Right, the surface is when the board is pulled out, rinsed, and acid remains behind in pores, concavities, etc. Etching rate in those pores need not be very much at all, after all it'll consume the etching capacity of that tiny amount of solution very rapidly; the problem is trapping salts in the pores of the metal, under soldermask, etc., to eventually absorb enough moisture and oxygen to cause breakout and continuing corrosion problems. Chlorides are pretty nasty with copper, it doesn't take much. And they don't evaporate on heating, the water does but there's no guarantee the deposit remains dry forever. (Hydrous cupric or ferric chlorides do decompose a bit on heating, but some chloride is always left.)
Probably, the etching is better controlled than in olden days, leaving a smoother surface with less porosity? A more thorough rinse/wash step might also be relevant. Say, using a neutral (non-oxidizing) complexing solution to dissolve residual copper chloride/oxide.
And yeah, if nothing else, might as well avoid them [acute angles] because they're ugly.

Tim