yep this is what I do, though I haven't collaborated in a bit. it really is the best workflow at this point and it can be as granular as you need/want it to be.

True, but this is how we used to do it in the old days, when track limitations forced us to bounce with all the outboard effects baked into the track. Producers might not physically ride the faders anymore, but a lot of them are very reluctant to even track vocals through compression onto disk. They want to keep everything available for processing later on.

Now, is this a good idea? There is a lot to be said about having an upfront plan about the usage of effects and placement of parts on the soundstage and committing. This might be a lost art, as I keep seeing Pro Tools files with hundreds of tracks, with stacks and stacks of effects.

I mean, as an older fart, I do think there's value in committing to an idea. Though, I could also see benefit in having the processing be malleable until the last possible second. I mean, imagine if you're working with an egomaniac artist who keeps being wishy-washy on which sound they like best. If the effects are being printed and they want to go back, it could become a huge pain in the ass. AT THE SAME TIME, I think there's merit in telling someone 'no' so the project can move forward. I dunno, I guess it depends. I can see it both ways.

Freezing/bouncing also lets you have more determinism.

Plugins with (often intentional) indeterminism in them, variations in OS and CPU architecture, variations in how fast their CPU is (a billion plugins on a track may run smoothly for one person and destroy someone else's machine), variations in sample rate, etc. can subtly or substantially affect how something sounds.

And in general it just makes things way easier to share even ignoring the issues of getting people to install the same plugin at the same version and the IP issues.

This. It’s the canonical way to collaborate with audio. I even ”collab” with myself because loading old projects is fragile in general. Always has been.