In this case, it's more of setting expectations for the player. People aren't dumb - they know in other games that houses are instanced, therefore if the house is not instance in GW2, something's gotta give. The expectation must be set in that if you're renting a spot in a neighborhood, you are renting it, not buying it forever. The baseline is still an instanced house, which the players would never lose if they take a break from the game, i.e. it follows the standard of other games. But the feature itself needs a stretch goal, otherwise it's buy a house, you're done, then log off because you ticked the checkmark off in your head. There's no engagement if there's no prestige and no ongoing investment.
EDIT: I also just wanted to add that I like the idea of a real estate market developing around specific zones being more desirable than others. First, it creates prestige around having beachfront property etc, and also creates cheap markets where people who don't want to pay much will just live in a grittier zone. But most importantly, it addresses a big problem with the open world in that many zones are never visited because there's simply no reason to. You can create a dynamic where zones now have neighborhoods and associated costs with them, breathing life into dead zones. This is really what ANet wants, a cost effective way to boost player engagement in old content while recycling as much as possible.