For example, the neighborhood of Mexico City where I live is called Santa Cruz Acalpixca, and the next neighborhood over is San Gregorio Atlapulco. In the State of Mexico--which borders Mexico City to the north, east, and west--there are towns such as San Juan Teotihuacan and San Mateo Atenco.
Follow-up: Is this pattern uniform throughout Spanish America? Is there a general trend? For example, I've seen:
The easy answer would be: it has to do with conversion to Christianity in early colonial times. And with patterns developed connecting town (and people's) names to Catholic teachings or to saints' names. The conversion of native Americans was central to Spanish colonialism and many techniques were connected to aid in conversion, especially by the religious orders: including alphabetization, burning of native codices, theatre... Adding Christian names to indigenous town or neighbourhood names can be seen in this light; but also as a symbolic act, since place names carry power.
Let's first look at some general patterns briefly, and then turn to some of those Christian town names.
I just mentioned how place names carry symbolic power. Well the Spanish (like later colonial powers) liked replacing native with European names as a sign of their rule. Look just at the first American Spanish settlements by Columbus in the Carribbean : the island Hispaniola (named for Spain, because of course) and early settlement Isabella (after the Spanish queen). Bartholomew Columbus called an early settlement La Nueva Isabela, after an earlier settlement in the north.In 1495 it was renamed with current name "Santo Domingo", in honor of Saint Dominic. So at the first Spanish settlements we find a naim for Spain, one for the queen and one for an important saint – symbolizing state and church authority.
The colonisation of th Spanish Carribbean meant the near eradication of its native population. And while Mexico/New Spain saw a similar demographic catastrophe, the Spaniards attempted to not repeat the large scale masscres of the Carribbean seen now as mistakes; plus there were important native empires they could build on.
So while (I think most) native place names in the Spanish Carribbean were replaced, this was not the case for Mexico. Instead we have different patterns there: many names esp. of larger towns and cities would remain, albeit often adapted to Spanish pronounciation – new-founded cities would often receive only Spanish names – while smaller towns or dependencies could have either of those or the mix that you describe. We'll get back to this in the end and now turn to some adoptions of Spanish names.
I couldn't find a study just for place names, so will draw one of native people's names, with many parallels with town names as we'll see: Rebecca Horne's Nahua Naming Patterns in Postconquest Central Mexico. I'll distinguish between names tied to Catholic lore and to religious orders/saints since they had seperate jurisdictions, although in reality the naming was less seperate than this.
Rebecca Horne notes how Nahua people with two names would often draw one of them from Catholic lore and doctrine: among them de la Cruz, de los Angeles, de los Reyes and Domingo Ramos (referring to the Cross, angels, the Biblical Magi and Palm Sunday). One of the most popular of those was de la Cruz, showing a Christcentric devotion. All these are also common for towns as well, supposedly with similar connotations.
We can note such patterns from the beginnings of colonisation. An example comes from Díaz de Castillo's “Historia de la conquista”, describing an early episode during the conquest campaign against the Aztecs. Cortés had brought religious images in order to convince indigenous people of their supposed idolatry from the start. When they came to Tabasco
the sacred image of Our Lady and the Cross were set up on the altar and we all paid reverence to them, and Padre Fray Bartolomé de Olmedo said Mass and all the Caciques and chiefs were present and we gave the name of Santa María de la Victoria to the town, and by this name the town of Tabasco is now called.
Directly after this the friar started preaching and baptizing townspeople (incl. The famous Malintze). For the founding of the first Spanish settlement Villa Rica de la Vera Cruz, modern-day Veracruz, that was needed to justify Cortés' taking control of the campaign, Díaz explains
we .. landed on 'Holy Friday of the Cross' and 'rich' because of what the gentleman said who approached Cortés and said to him: 'Behold rich lands! May you know how to govern them well.
These examples show how a town's name could be replaced or chosen for a new settlement (like with later Spanish foundings like Puebla) – which shows some patterns but is still different from those of your question. In addition, the name María was also very popular for both towns as well as Spanish and native women – it surely reflects the beginnings of the Marian cults that would become more important in the 17th and esp. the 18th c. To this can be connected terms tied to Marían devotion like de la Concepción, de la Visitación, de la Asunción, and could also be tied to certain feast days. So far we've seen place names tied to central tenants of Catholic teachings like the cross or the virgin Mary, but there are of course many more place names.
Following the mass baptisms led by the religious orders in central Mexico, Christian saints' names became widespread by the 1550s. Parish priests would encourage Nahuas to take up the Spanish practice of naming a child for the saint connected to their birth or baptismal date. There are parallels here with pre-hispanic customs where calendrical names could influence on the choosing of birth dates, making this practice easier to adopt for Nahua.
Certain names prominence seems to suggest their popularity
among both Nahuas and local Spanish clerics. The names Juan and Juana reflect the patron saint of the local sociopolitical unit [in Coyoacan], San Juan Bautista. Similarly, the popularity of the given name Dominga, … and Domingo undoubtedly reflects the fact that Coyoacan was within the doctrinal domain of the Dominican order. [This pattern] was a common practice throughout colonial central Mexico, conspicuous in late 16th c. Culhuacan and the late 17th and 18th c Toluca Valley. One suspects that naming after resident religious orders was also a common practice, Francisca and Francisco and Agustina and Agustín being esp. popular names in Franciscan and Augustinian areas... [Horne, 111]
The name Juan referring to St. John the baptist was probably the most common saint's name among Nahua. The Oaxaca region still has many towns called “San Juan Bautista [other name]”, showing the influence of a regional saint's name on intertwined naming practices of people and towns alike. An important point here is the parallel between people's and town's names: where people would often take up the patron saint's name tied to their place of origin, making it difficult to distinguish between both naming processes. With towns I could also their patron saint being connected to a specific date - like the town's festival, the date of Spanish contact and/or conquest, even possibly a pre-hispanic feast date etc.
We can note here both native and Spanish influnces. Nahuas may have preferred names connected to their home towns and local saints as expressions of traditional micropatriotism (identity being tied up with home states or altepetl), while clerics would have encouraged this use as a recognition of the local religious orders. What is more, mendicant names retaining the “de San” part or others taken from famous friars (like Pedro de Gante) were seen as signs of belonging to the Nahua elite. For towns the “de San” formulation would have been natural, to designate it as connected to a specific saint or order.
So far I've only looked at one part of your question. I haven't found an explanation in the literature for this mixing of Christian with indigenous names (yet), so will add some educated guesses. Horne mentions the common mixing of both names for people, but says that by the 17th most Nahua held only Spanish names, so different from towns. As mentioned above major indigenous cities usually kept their names, such as Tlaxcalla, Tacuba, Xochimilco and Tezcoco – most of those held city status. The mixed names were then more confined to smaller towns or dependencies, or even neighbourhoods of such larger indigenous cities. Some of your examples from Mexico City were actually the four indigenous neighbourhoods of that city, all prefaced by a saint name.
A reason I could see for this mixing is the major importance of the ethnic city state or altepetl throughout colonial times. Briefly put, in pre-hispanic times people had identified more with such local political units than with, say larger entitites like the Aztec alliance. This importance of altepetl continued with modifications throughout the colony. I'd argue that since the Spanish genereally attempted to build on indigenous political structures like the altepetl, replacing altepetl (or town) names would have been unacceptable. It would have represented a break with these continuing strong local identifications (although note that such brutal breaks did happen e.g with the congregaciones).
In conclusion, I'd then see this mixing of existing native place names with Christian ones as a compromise: where local identifications could stay largely intact in towns, while connecting them symbolically to Catholicism. This would also fit with larger patterns of cultural transformation in Mexico, were the native faith or other traditions were not simply replaced with European ones (despite strong attempts), but rather indigenous elements combine with them to produce different forms.
I'm sure there were many more adaptations of names in post-colonial times that I can't comment on, but hope this already gives a good idea.