I don't really think LSP 2.0 and Vixen 3.x are going to be in the same ballpark in terms of features and capabilities. At least not in 2011. THat said, I'm still very excited to see what Vixen 3.0 can do!
I participated in the LSP 2.0 beta testing and was really enjoying it by the end. David worked his tail off fixing everything we could find and then some. It truly is powerful and capable. You will not find anything else which can talk to all the hardware and protocols like it can. Learning it requires a lot of time and patience. I still don't understand a lot of what it can do with it, buts its been fun discovering that. I find myself being productive with it and not fighting or complaining about it constantly. I'm sure there are bugs left, probably quite a few ... but the really big annoying ones seem fixed for the most part. The amount of good far exceeds the bad at this point. The underlying performance issues from 1.8 seem fixed. I even pushed to get the chase to operate in a better way and to operate on virtual controllers or layers and that is extremely powerful now. You can do everything you can do in S3 and probably then some. It's probably quite a bit more advanced than even S3 w/superstar, esp. when it comes to RGB object oriented programming.
I bought LSP 1.8 and found it basically unusable and extremely frustrating, slow and unreliable. LSP 2.0 has made some very nice improvements ... and (LSP) David is trying to make a living, pay the bills, seriously. I think some of the LSP 2.0 pricing decisions are basically required to keep it alive and viable ... and its features and capabilities are comparable to Madrix and SuperStar Lights ... but better than both.