What's the difference between Flatten Comp and Clip Bounce?

Depends what you want to do.
What do you want to do?
 
In my world, I've never used flatten comp but this is a tool that you use when you're working with take lanes which are comp'ed together to create a single track from your takes.
 
Bounce to clip is what it says - Select a clip or a range of clips and they will be combined into a single composite clip - this works for Mid and Audio.
I don't think flatten comp works on midi but not having used it (yet) I can't say for sure
 
When dealing with audio clips, bounce to clip will remove  Audiosnap data and Melodyne data when bouncing

Late to this party but I have to ask. In both cases, I gather they re-write the data into that new track / comp. Meaning, however you set bounce to (24, 32 or 64 bit) is the quality it is rendered at. I remember years ago, I'd learned Pro Tools let you combine clips, without rendering. Meaning, you put 5 edited clips together into what looks like a continuous "clip" like you'd bounced it. Yet, it's NOT rendered. It's the old five, but melded together. Meaning all the joys of easily moving it around or whatever (it has its benefits). Rendering in Sonar (bouncing, flattening if it means the same process, as I think it does) leaves the conundrum. Do you set it to 64 to get what is nearly identical in sound as the unprocessed originals (creating a massive file vs. the original 24 bit) or keep it at 24 bit for no size increase (and live with the fact that you lost quality. I've opted historically only to bounce in rare cases when it was so beneficial that any loss didn't matter. The "joining" vs. rendering feature is something Sonar has never had that I could really use for me. Just being idealistic.