I agree with you guys about the facts that there is no specific “good number” in the number of centers, multi-center clinical trials (MCCT) are expensive, there are a lot of aspects involving in MCCT. There is an article named “A guide to organizing a multicenter clinical trial.” on PubMed that showed a strategy tackling MCCT. Like many have mentioned, the success of the trial majorly depends on the fluidity of collaboration between all collaborators toward a common goal.
I think the paper mentioned many good points in assuring the fluidity of communication. For example, there are various useful consensus building methods such as focus group discussion, nominal group technique and the Delphi method. These methods will help gather and represent the knowledge and experience of all participants.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2917608/