Let me give you some points about that "hard to implement":
The engine already supports individual instances that scale and allow for visitors (neither you or me do know the limit of how much of those can be active at the same time though, but that limit can be stretched by placing a few of those doors in different areas). So creating one door leading to multiple player houses of a certain type/design is utterly possible without much dev effort.
The engine already allows for placement of structures to nodes that can be defined from different node types. So making the "housing" a basic design with customizeable notes is utterly possible without much effort now.
The engine keeps track of completed quests already. So allowing for a trophy wall in that housing is utterly possible right away.
The basic designs for different cultures are already there. Even sample rooms and houses are already found in game ready to use for "housing" with a simple copy/paste + cleanup + node adding.
The engine also already supports doors closed till certain quests are done or active (thus allowing to place doors in the housing leading to areas that need to be earned or bought to expand).
Also you might be quite wrong about that "small subset of the population". First, it might be small because there is no housing ingame and the guild city building is extremely tedious, thus putting the kind of players who would like it busy doing other things. Second, it might not stay a small subset, if the housing is done right. Third, it might not even be a small subset right now and the "impression" or "evaluation" is wrong.
There is a reason why games allowing for modding or player customized content (nwn, sims, ultima online) were (and are) popular even when the tech and graphics were outdated.