Should I try to fix printed clipping on tracks before doing time stretching?

That is a good question. There are several different methods to perform time stretching do you know which cake employs? If they are using sinusoidal spectral modeling my guess would be you will see better results if you get rid of the clips first. I really can't say how many clipped samples in a row would cause significant distortion in the modeling. Or if my assumption is the correct one. With no experience to go on my gut says get rid of clipped samples first. Interested to see what others have to say. 

Yeah, I'm thinking it would probably be better to clear up the tracks first however it's kind of an odd situation. I'll explain a bit further.
 
I'm working with two takes of each song. One of those two takes will end up being the foundation. Basically the the best performance with the least tempo wonkiness will be chosen.
 
There are very few parts where clipping occurs and they are pretty brief but they are noticeable so it has to get fixed. These clips do not necessarily occur on both of the takes so I will be trying to simply swipe parts from the less desirable take if I can to replace where it clips on the take I'm using. On some of the songs though the clips are in the same spot on both (loud screams).
 
I also have a completely OTHER set of takes that were recorded at a different time location where the vox were isolated but the performance isn't as energetic, they are quite a bit slower AND they'll require an export from my old system. I'm still considering either using these takes in part or in whole but I haven't decided.
 
Considering that it's only going to be very small segments that need to be corrected, I haven't really put all the tracks under the microscope yet (I'm building up my familiarity with the tracks as I work with them) and that I'm not sure which method I'll be using for fixing the clips (one of which is the time limited Izotope RX demo so I'd want to do it all before the time limit expires) it really is one of those situations where I think I could spend WAY more time than necessary. Basically if I start with correcting the clipping now I'll have to go through all of the vocal tracks instead of just doing the parts that really need it.
 
Make sense? One other thing I was thinking was seeing as how I'm obviously cloning the tracks before doing the time stretching that if it does get screwy I could simply take copies of the unstretched audio, correct the clipping and then stretch those to match the other parts I've already done. That however would mean my transient points where I'm stretching won't be the same as where I stretched the rest of the tracks.
 
I know... I'm weird. Anyway, I'm currently starting with the tracks that don't have clipping or their second take has clean parts for replacement so I've got time to mull this over. Cheers.

Well that's the thing. It's all kind of going to be trial and error as far as what takes are gonna get used and I was hoping to save that for the actual post time correction editing phase. It could end up being that I only need to fix one or two very short sections at that point. If I do it now I'd have to do 9 songs worth x2 for each version of the song and then x2 again because there are two vocal tracks. Just trying to save some time.
 
Thanks. I'll just need to think on this a little more. Got a bunch of meatworld mayhem that just popped up too so that's throwing a wrench in the works. Ugh.