There has not been any definitve comment by the LimeWire crew on that topic yet. It will depend on the specs of gnutella2, that are (according to Shareaza Mike) going to be released once Shareaza 1.7 final is released. So far, Gnutella2 is a closed protocol and you can only make guesses on how good or bad it is.

The whole gnutella2 issue caused a lot of discussion on the GDF. First, Shareaza implements a new protocol without even giving the other developers the chance to comment on that or publicly releasing it. - Then they call it gnutella2, which might lead people to the assumption that this was indeed the 2.0 version of the gnutella protocol. (Actually a project called gnutella2 existed long before Shareaza picked the name up: http://sourceforge.net/projects/gnutella2 ).

I suspect that LimeWire will not implement Gnutella2.

Level one: The new messages will require a lot of rewriting, testing and so on and if I were a developer I'd think twice before implementing two sets of network messages for basically the same tasks instead of just adding a few additional messages maintaining the basic structure of regular gnutella messages (the way BearShare did it).

Level two: might be an approach to use UDP for data transfer. If they want to use it only for messages it would be covered by GUESS. If they wanted however to use it for file transfers or streaming media I would not consider that a good idea at all.

Level three: seems to be one part GUESS and another part CHORD.

And as for level four - I could only make vague assumptions about what that should mean.