Cities such as the Khmer Empire's [Angkor Thom] (https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcTLT5YzifRl3aZJknKwINjsm85t8sTWKUIdr6tkagRnk9eqX2thmm0uvEZJhw) seem to indicate a large canal based city. As do the Majapahit Capital of [Trowulan] (https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcTV8aYan3EjArocaZfwtz_n1gMCF0XA-2-MNjDuBZxDnaQSP05eM2qi8E7F). However, contrary to popular depictions of Medieval and Early-Modern cities in Europe, the Middle East, and China as dense and walled in and/well defended settlements, this does not seem to be the case with Angkor Thom and Trowulan. [Illustrations] (https://encrypted-tbn0.gstatic.com/images?q=tbn:ANd9GcQ4Ac61Svanx65q8qK3a69r03ABkppiYmEJHJf6zCHwvKu1gAlBS9qnxAu4) of Trowulan seem to depict a city thats closer to a collection of government buildings than settlements such as Paris, Baghdad, or even Tenochtitlan. If there exists fortifications around these Capitals of great and powerful Empires, why are there no ruins of these fortifications that would most likely have been built by the same materials as the ruins of the surviving temples. Do these two cities function and ceremonial capitals similar to Persepolis in Persia? Are there any records and descriptions of what ordinary cities would look like and how they were defended?
Early Modern Southeast Asian cities appear to have been usually comparable to their European counterparts in size. In 1650, there were probably around four or five cities with populations exceeding 100,000 (Thăng Long in Vietnam, Ayutthaya in Thailand, Ava in Myanmar, Makassar in Indonesia, and perhaps another Indonesian city, either Aceh or a Javanese one). In 1600, there were only around five such cities in Europe too outside Italy (Constantinople, Paris, London, Lisbon, Seville).
If anything, the area was likely more urbanized than Europe. Southeast Asia was very thinly peopled -- as the king of the Indian realm of Golconda once said, "It is true, I admit, that [the kingdom of Siam/Thailand] is of greater extent than mine, but you must agree that the King of Golconda rules over men, while the King of Siam rules over only forests and mosquitoes" -- and this meant that urban populations took up a significantly larger proportion of the population compared to Europe or (especially) India.
Unfortunately, insufficient archaeological data, coupled with overreliance on European accounts that have a tendency to exaggerate city sizes, mean that the size of most urban populations are disputable and population estimates often differ by an order of magnitude.
Just consider the case of Melaka. Melaka was the most important Malay city in the fifteenth century until its fall at Portuguese hands in 1511. As its conquerors reported, it has "no equal in the world... a city that was made for merchandise, fitter than any other in the world." Or as Malay accounts say: "From below the wind [i.e. Southeast Asia] to above the wind [i.e. outside Southeast Asia] Melaka became famous as a very great city, the raja of which was sprung from the line of Sultan Iskandar Zul-karnain [Alexander the Great]; so much so that princes from all countries came to present themselves before [him]. So what we do know is that Melaka was a biggish city, famed for trade and opulence.
What we don't know is the population. Population estimates for Melaka (a big modern city which has been the focus of little archaeological attention) range from 10,000 (equivalent to a smallish European city in 1500) to 200,000 (equivalent to the single biggest sixteenth-century European city). On the other hand, Melaka's population almost certainly fluctuated significantly according to the monsoon seasons that trade was dependent on.
Until we have more archaeological information and more sources (especially indigenous bureaucratic ones, as rare as they are), the real nature of most Early Modern Southeast Asian cities will remain obscure.
Finally, there was no archetypal "Southeast Asian city." After all, the urban centers of Early Modern Southeast Asia featured five different religious traditions: Islam, Theravada Buddhism, Confucianism, Christianity, and local animist beliefs. The great cities of the region belonged to four major language families, each of them as unrelated to each other as English is to Finnish: Tai-Kadai, Austronesian, Austroasiatic, and Sino-Tibetan. What was true about Melaka, a pure mercantile port with little of an agricultural hinterland, was not true about Ayutthaya, a flourishing port whose major exports included rice, and what was true about Ayutthaya was not true about the inland riverine capitals of Ava.
This alone wouldn't be a satisfactory answer, though. Below, I'll try to address one of the two specific examples you mention, Trowulan, then look specifically at the case of Ayutthaya, a fairly well-described and among the largest cities in Early Modern Southeast Asia.
As late as 1815, there were only five cities in Java whose populations exceeded 20,000, and their inhabitants put together represented less than 7% of the island's several million people. Three of these were colonial ports ruled directly by the Dutch (Batavia, Semarang, Surabaya).
These small urban populations appear to have been reflective of most of the precolonial past. These are the population sizes that the Portuguese writer from the 1510s Toma Pires raises:
In sum, with the possible exceptions of Demak (the only case where Pires references hearsay) and Surabaya (which discusses "fighting men," not urban populations), there was no sixteenth-century Javanese port with more than seven thousand people. Coastal Java, it appears, was a very non-urban society.
This pattern prevailed for most of Java before the initiation of penetrative colonial rule in the nineteenth century. Population densities slowly increased in most of the island through agricultural intensification, but urban populations remained small and fluctuating.
The biggest reason for this is the settlement patterns of rural Java. Javanese rural communities were fixed in a certain region and were often so stable that many communities mentioned in tenth-century land grants are still flourishing in 2018 in the very same region and with the very same name that they had a thousand years ago. But when population densities increased, instead of expanding into something bigger we might call a town, the expanded village simply underwent a form of mitosis into separate but interrelated settlements in the same region, often clearing new land to do so. The number of settlements in rural Java increased significantly, but each settlement's size did not grow very much, at least not enough to become an urban settlement. This was the case in Java fairly recently, into early colonial rule. Economic structures were similarly decentralized, with markets moving on a five-day basis rather than staying in one market town which might have emerged into a permanent commercial center.
So there were few cities in Java under Majapahit. Were there none? This brings us back to your original point:
Illustrations of Trowulan seem to depict a city thats closer to a collection of government buildings than settlements such as Paris, Baghdad, or even Tenochtitlan.
That picture shows the scholarly consensus up to 1991.
Up to the mid-1990s, historian Jan Christie's assertion that "the later, fourteenth century site of the Majapahit capital... shows few urban characteristics" was widely accepted. Then, in 1991-1993, the Indonesian Field School of Archaeology held an archaeological survey of the general Trowulan area. It surprisingly showed that far from being a purely ceremonial capital surrounded by rural hinterland, the Troluwan monumental complex was merely the center of around ten thousand hectares of dense populations, intensive trade, and specialized manufacturing -- about the size of San Francisco. Subsequent research has supported the findings.
To be sure, much of the area must have been devoted to reservoirs, canals, mansions, and gardens crowded with what one poem calls "the most excellent of flowers." Still, the densely populated area around Trowulan was more than twice the size of eighteenth-century Javanese cities like Yogyakarta. If we assume that population densities in Trowulan were similar to these eighteenth-century cities, the Majapahit capital must have been home to 200,000 people or more at its fourteenth-century height. This was comparable to Paris, the largest European city of the century.
Trowulan also appears to have flourished as an urban center in the fifteenth century, even as Majapahit was undergoing economic decline. This suggests that the city had economic viability somewhat independent from its status as a political center, even if the city could not ultimately survive into the Islamic period. After all, as one Majapahit poem mentions, "[e]ight thousand cash [Chinese copper coins] every day from the markets is the share" of the official in charge of protecting the city's commerce.
From what little we can gleam from literary and historical sources (temple reliefs make very little mention of urbanism while elite writings are preoccupied almost solely with elite civic and religious life), much of the city appears to have been organized into kuwu, discontinuous residential compounds inhabited by a nobleman and thousands of servants and artisans, the professional clients of the noble. Still, we know very little about how they functioned and will not without more archaeological research. How large was the average kuwu? How large the spaces between them, and what were they used for? Were there rice paddies between the kuwu? A Majapahit source asks: "What is called the nagara? All where one can go out (of his compound) without passing through paddy fields." The Javanese word nagara comes from the Sanskrit word nágara, "city," but was often used to mean just the royal palace compound. Which is it here? There is still a lot more archaeology to be done in Trowulan.
[Continued]