It is a given that when major changes are implemented in software that there are going to be problems. With thousands of people running any number of soft and hard configurations it's going to take awhile to sort out bugs and other problems related to implementing a protocol enhancement such as ultrapeer.
Rather than bitch about the fact that I haven't yet seen the advantage of the ultrapeer implementation I will level a few comments.
1) most gnutella users use limewire to get and share files. My experience is that the 2.0+ versions of limewire simply don't return as many hits for a given search as the older versions did.
2) I went looking for information on what was this new ultrapeer enhancement. There is nothing describing it's implementation or what its intended purpose is in any of the common FAQ or user documentation. If a programming group makes a signifigent change in a product like ultrapeer as an example, where the function of the software is as markedly changed as 2.0+ is, it would be to the benefit of the community as a whole to provide some information to the users about what the change is and why, as well as how it works.

If the implementation of the ultrapeer method is a step in the chain to making the gnutella net function better then thats great. But without some information as to the why and how this new connect method is superior to earlier implementations large numbers of folks are simply going to go find something they think functions properly.

Lets see some info on the limewire website about ultrapeer, how it works, what it is, and why it's better than previous implementations. The general public aren't programmers, they just want to fire up their client software, make a search and grab what their looking for. Frankly 2.0+ is disappointing in it's functionality, and without some clarification about what the new changes are suppose to improve, Limewire will only loose marketshare....