In a recent debugging session with one of my users on a very busy server, we only discovered that we needed to tune MaxStatups after increasing the LogLevel setting to VERBOSE. I would think that this kind of condition warrants reporting at the default LogLevel (INFO). It would have shaved a few days off of the debugging process in our case if this had been the case. Thank you for your consideration in this matter.
What version did you see that with? There was a change to improve the logging that shipped in 8.4 (bug#3055, commithttps://github.com/openssh/openssh-portable/commit/9c5f64b6c)
Yes, there is ample MaxStartups logging at info/error level now, specifically whenever MaxStartups starts or stops throttling and periodically while in this condition. see https://github.com/openssh/openssh-portable/commit/9c5f64b6c
closing resolved bugs as of 8.6p1 release