I don't know why it has an effect on those who have applied the 8.3 tweak, because earlier updates with those that had the same 8.3 tweak applied didn't have this problem.
I don't know what has changed, but what I do know the amount of people who have applied this 8.3 tweak is going to be infinitesimal by comparison the the 10s of millions of avast users and those that don't have this 8.3 tweak applied. So I can understand where they are coming from when they say they suggest reversing the tweak and even give the .reg to make it easier.
I have never felt it necessary to apply this tweak to stop the conversion to 8.3 short name conversion. There might have been a performance issue when this tweak was first thought of but I doubt that now as many have reported there either appears to be no noticeable impact on performance or none at all. I would tend to believe Igor's comment in this post
http://forum.avast.com/index.php?topic=28102.msg230206#msg230206 about the Short File Name generation (not its minimal impact). When the impact on the tens of millions not having any problems effectively being forced into another program update IMHO is much greater. These kind of decisions are taken all the time by all program developers (it is called change control, the impact of your changes Vs the impact of not making an immediate change) MS is a good/bad example. It is just that you don't see them being open an honest and up front about it.
Would abandon avast over this if I had previously applied the tweak, absolutely not, I would apply the .reg reversal on the advice of people who are working at this sort of system level application programming that there really is no performance impact.
Not when you consider the activity of the application programmers in these forums directly responding to problems I have never seen this kind of input by the software developers/programmers. However, it is your system and your choice but you should consider your own personal experience and the support levels with avast.