Valid point. Atlanta has roughly quadrupled its population in the last 30 years -- in the '80s and early '90s, the growth was concentrated in the suburbs. Starting with the run-up to the Olympics, a lot of development has gone on downtown, reversing white flight to a degree, because the traffic in the 'burbs got so nightmarish.
Intown housing has been on a years-long boom, so far (knock wood) with enough population growth to support it. Developers have moved eagerly into neighborhoods formerly on the skids, both with new construction and a lot of warehouse and office conversions to residential. In designated enterprise zones, developers get property tax abatement -- but a condition of that is that they have to set aside X number of units for rents below $Y a month. I don't know if they're required to accept Section 8 or not.
Whether poor residents pull a neighborhood down or the neighborhood helps pull them up is in part a function of numbers. And there Atlanta also benefits from local trends. The new housing -- in which I include renovations of neglected properties as well as construction or conversion of new ones -- has been sufficiently widespread that the population of section-8 recipients in a given neighborhood doesn't hit the ripping point (an overused, but still useful, concept).
Of course, what I'm talking about is only a little more than a decade's experience. It could still go wrong, especially if Atlanta's boom slows down (if, for example, we have to halt new development because we run out of *&^ing water). And that's where your point about one-size-fits-all programs hits the nail on the head -- the programs need to be flexible enough to adjust not just from place to place, but over time.
In cities where population growth is stagnant or even negative, where there's aging surplus housing, the dynamic would be very different. Even leaving out medium-sized cities like Camden, NJ and Gary, IN, which are just basket cases, and Detroit, which is headed toward becoming a medium-sized city,