Tyler Cowen at Marginal Revolution laments zoning, though he speculates on the cases in which it would still be preferable. I have written before on my admiration for Fischel's Homevoter Hypothesis, which is the basis for two points: 1) Despite its costs zoning may be a net positive on the whole and 2) even if 1 is wrong, it is not so deep in the red that it should steal focus away from libertarians on other more pressing concerns.
In short, zoning gives us a near vertical supply curve in housing, which is already more stock than flow, hence deadweight loss is likely to be extremely small. Caplan (2001; Public Choice) and Powell (2006; Public Choice) have been overly pessimistic about the role of capitalization and taxes, viewing it as a means for local government exploitation because, in their view, once taxes are raised property values fall and there is no sense in "voting with your feet." That is just the point though, at the local level voters feel the pain and gain of their politicians actions. Since property values represent a long time-horizon, even short-sighted politicians must think long-run to get reelected.
Why is rent control not more prevalent? My guess is if you survey people they would tell you that rent control is good. They abandon this belief though when their own property values are on the line. NIMBYs can only locate in municipalities if they provide enough compensation to the existing residents to offset any depreciation of housing values. The EPA need not worry itself about locally polluting firms, as they will have to negotiate a compensation package with a municipality, and attempting to control that firm's waste will both be unnecessary and will undermine voluntary transactions between communities and firms.