This is a great question- and brings into focus the difference between imperial projection of unity vs the actual situation. Though the details I'll give are based on the Ottoman Empire, the lessons behind it are, I would argue, applicable to any such empire (even, though of course in very different ways, to modern nation states).

So the Ottoman case- roughly speaking, the Ottomans can be considered more centralised than their mediaeval/early modern European counterparts because they were primarily an imperial system, and as such did not carry the same feudal inheritance of the remnants of the Western Roman Empire that emerged from the combination of legally enforced guilds, the local power of Dux, later dukes, and the spread of Catholic Christianity. Instead, the state tradition the Ottomans inherited was a combination of the Central Asian Turco-Mongol tradition, the Sassanid Persian state tradition, and the Byzantine imperial tradition. All of these of course have aspects that are arguably comparable to a feudal state, but at their base are geared towards imperial structure. This imperial structure differentiates from a feudal structure in part due to the centrality of the imperial throne- the empire is not a highest among peers, and is not bound to custom per se, but rules from an imperial Centre that is clearly above the periphery and can legislate for even them (hypothetically)

Something to note about empires, however, is that thet are not necessarily centralised. Many empires, including the Romans, the Chinese dynasties, the British, are distinguished by their inclusion of a variety of peoples, as opposed to one "nation" or "proto-nation". What this oftentimes means is that the customs, religions, and state traditions of these peoples also get plugged into the imperial structure- making the empire a tapestry of Interwoven power structures.

So what about the Ottomans specifically? Drawing from İnalcık's Classical Age and Tezcan's Second Ottoman Empire, we need to split the empire into rough phases: the Beylik phase, the Kul phase, the Mevali phase, and the Tanzimat.

During the first phase, we can't really talk about a centralised or imperial Ottoman state. This is the period where one could argue the Ottomans were most similar to a feudal state in terms of the division and centralisation of power. However, instead of castles and cropland, it was tribal affiliation that mattered- the Ottoman bey was the premiere tribal head, but did not have full control over the internal dealings of the other beys. They often used this to their advantage, placing warlike tribes immigrating from Central Asia, Iran and Anatolia onto their frontier with Europe, where they would make conquests to include their tribal land, which was only through their ownership under the umbrella of the Ottoman state.

The conquest of Constantinople brought on the second phase. With the prestige this conquest brought, Sultan Mehmed II was able to put the other beys in line, drawing from Byzantine tradition to codify the secular "kanun". The devşirme system went into full swing here, with higher ranks and regional government posts being taken from hereditary tribal chiefs and given to centrally trained slave lords, who were disallowed from holding office for too long. This is the source of much contemporary European political writing. Especially in the Renaissance, thinkers re-discovering Greek philosophy drew comparisons between the Persian/Greek East-West duality with the Ottomans and Christendom. Ottomans were seen as a totalitarian, tyrannical system, to be compared with the growing understanding of rational statecraft in Europe at the time. The Kadi justice system and the sultanic ownership of all land is contrasted to law and private property that was meant to hold power over even the sovereign. These political themes are also why the Ottomans were displayed as excessively centralised- they were already the great evil, so associating absolute monarchy and imperium with them was a way to further promote Constitutional monarchism and even republicanism.

The third stage actually sees this trend occur in the Ottoman Empire as well, or at least this is what Baki Tezcan argues. He writes that the rapid succession of multiple sultans, their inability to command outside of a coalition with other power brokers, and the supplanting of sovereign "kanun" (secular law) by Islamic shariah, were all signs of the conscious limiting of monarchical power in favour of a more inclusive constitutionalist trend, led by Janissaries involved in trade and ulema (religious scholars) who were the sole interpeters of religious law. As such, both military and legal power grew more and more distant from sultanic hands. It is in this period that we see the obsolescence of the devşirme system, viziers and governors hiring roving mercenaries as personal retinues, and local negotiation of power gaining more importance.

This local-systematisation of power was, however, very Muslim-focussed. State officials, military, and of course religious officials were almost always Muslim. Kanun had originally been a means of modifying and codifying local customs, which left more room for non-Muslim participation in the state. With the ascendancy of shariah, this became less the case. The eventual drifting away and then revolt of many non-Muslim communities was in part due to this. Combined with the politicisation of the military, which made it almost impossible to reform or form new armies for the changing military systems of the time, pushed officials under the grand vizier to launch the Tanzimat. It was an attempt at centralisation, but one which wanted, at least at first, to be less politically inclusive. It was also, however, subject to struggle between constitutional and absolute monarchists.

In short, the characteristic features of the Ottoman state during the traditionally labelled period of their "rise" was imperial and centralised, at least moreso than acl contemporary European feudal state. However, this centralisation had ongoing exceptions, and was also based on an absolutism- which was challenged in later periods. As such, we need to look at more than palace accounts, but also to political economy and legal history when asking whether the diminution of that absolutism was also a decline in centralisation. I would argue it wasn't, as even the great ayans in the 18th century all had similar, central, markedly "Ottoman" and not so much "local" upbringings.

TL;DR: Comparatively, Ottomans had a more centralised imperial system until the 17th century. After that, challenges to absolutism created rifts in the old pattern of centralisation, but didn't remove it outright.