In my experience with protocols, it often goes back to whatever the first person in a particular lab decided to do - not very satisfying scientifically, but at least it's consistent. It also depends on what you're most interested in - e.g. the cleanest possible supernatant, or the least damage to some component of the pellet. In general, I would look for papers on samples the most similar to yours, and then maybe test a few different times and speeds across the range they used.

Thank you barbara

Forgot to add - write to the people, if you find a promising method. Most likely they'll be flattered to be asked, and the worst they can do is ignore you.

Sure Barbara. Thank you