If I understood well, only transactions that are made while hard fork (around 12h pm gmt on August 1st) could never be confirmed. Once forks are separated, everything should be fine again in each fork.
If there is an hardfork, transactions could be nicely separated which makes it clear without messed it up.
If I'm right, only when there is a chain split (which is not hard fork), things could be tricky because at some point one of two branch will just be wipe out by the other which has achieve more computing power/ has the greatest difficulty of POW
That was the risk of the UASF BIP148 idea. If UASF BIP148 took traction, this would mean that the legacy chain would be wiped out forever, and only BIP148 UASF chain would survive and that would become de-facto BTC.
The good side is that you avoid the possibility of 2 bitcoins forever, the bad side is that if you are on the legacy chain and you sold your 148 coins for BTC you risk lossing it all.
Luckily, BIP148 is now out of the picture. I wouldn't worry about BCC at all.
If you want to worry, worry about november's supposed hardfork to segwit2x... now that 's something I don't see clear at all. I hope the NYA camp gets convinced of the mistake of hardforking in 3 months.
