i've used x1 on several occassions. my setup consists of about 6 synth tracks that i switch between via changing my keyboard controller's output channel. i'll have a couple of audio tracks my collaborator writes. live fx as well. also, matrix cells for triggering special fx. i open all the song cwps first and test midi in and out with them backstage, then all i have to do is carry my laptop and controllers onstage and plug in my audio interface. i've been re-creating the sets in x2 and not many a glitch that could not be figured out. most sets are about 45 min. the only problem i had was when i was practicing with a guitarist one time and my system bugged on me, so i rebooted. he broke a string and changing it took longer than it took me to get running again.
the most recent incarnation i'm using sonar live in is a three computer setup, all networked. sonar and pure data runs on one, with several vsts including reaktor to process and send osc controls. the other is a mac with pure data set up and a couple of audio and midi tracks in garage band. the third is an old win 7 machine with reason and pure data on it.
each computer uses pure data to read and analyze webcam data (the machines are in separate rooms tracking different people/objects) and send it to the other machines via rtpmidi, where those pure data (or reaktor) programs send interpreted info to one or both of the other machines. (note- this is for an art installation a friend and i are working on - not a concert). we had it running at a party recently for about 3 hours without a hitch in any of the systems. sonar was recording midi the whole time, and all the pure data components were writing osc parameters to script files.
i've never done any of this without at least a five hour test run in one sitting beforehand, though. if you know what your show is going to be like - do a dry run. i got some feedback in the forum that basically said "DON'T DO IT!!! ACK!!!" but i decided to try it for myself - it was the only way i would know.