before fork, client did start and shutdown much faster than the previous versions...
why rollback when everyone shouldve moved forward
There was some misinformation in the first minutes. v0.7 and prior had the bug, but a purposely configured v0.8 allowed it to be exposed and cause the fork.
v0.8 works fine for a user and merchant. Until there is a v0.8.1 then if you are a mining pool (including p2pool) or large miner (i.e., mining solo) then you want to use v0.7 as blocks mined using v0.8 could still produce blocks that v0.7 (now used for the longest chain) won't accept.
As far as your v0.8 client not syncing, that should be resolved by now. Has it?