The case for bouncing to audio before mixing

If you use a lot of Virtual Instruments (VIs) like I do (or hardware synths for that matter), how do you deal with the situation if you are called to do a re-mix or edit a year or two after you've already finished the piece?

What would happen if you'd upgraded your computer and the VI you used for that piece doesn't work with your new OS and you were called to re-call that piece and submit it for a high end advertising client? Would you be sweating bullets trying to recreate that awesome synth sound that the now obsolete VI could only create?

Why not take an extra few minutes and bounce, or render all your VI and hardware synth tracks to audio before you mix?

I know, I know, it takes extra time, but it really shouldn't be that big of a deal because you have a template set up already, right? :)

If you have all your outputs from your VIs already routed to audio tracks in your DAW (for instance assigning the outputs from the Multi page in Omnisphere or the Outputs tab in Kontakt), the it shouldn't be much more than playing the piece back once or twice, depending on the strength of your system, and recording the outputs as audio tracks.

Another benefit of this approach is to free up resources from your computer that it would normally have to use to process the VIs and direct those resources to all the cool plug-ins that you have but can't use because your CPU is maxed out from running all those VIs!!

If you find that you would like to tweak a VI part in MID again, just re-load the VI, edit the part and then re-bounce. Since what we do with audio on computers is a constant balance between computer power and our creative ability to stress our computers anyway, this work flow is a good compromise, IMO.

Any thoughts and comments welcome.