I might have made them sound better than they really are. With good I was more referring to how good it felt when actually getting a kill or two from a pretty dumb/cheesy place rather than overpowered places as the roof-exploit.
As an example. one we sometimes use is on top of the pipe to the left in the room @2:40. Only good for killing CT's coming from the cat walk entrance, because it is easily detected when coming from the t-spawn lobby.
Will that one be seen as an interesting dynamic the players are creating from your level's mechanics or will they be shut down? Hypothetical, if players find access to a certain room above, perhaps, some toilets, will that be fixed? ;P
In a patch like the one you mentioned, which of shown boosts are affected (in the video MrSkellyBones linked)? Only those that looks glitchy (like the one in the metal detector @0:58 when you got models clipping in your face) or boosts in general which you perhaps did not intend to create and don't like. The boost in the video @2:00 feels like a legit boost until you crawl out on the pipe and once again get metal meshes in the players' face @2:13.
When playtesting the maps, do you turn on team player collision to test boosts?
Edit: Wondering, because I'm currently between the sketch and blockup phase of a map myself and got Curious