No clear idea about current server world per se.
But I was running a computerchess account myself way back in time, and I am near 100% sure that my feedback will be to the point.
I was running the account for Gerrit Reubold, the developper of Bringer on the then hot server, so I was running an account that everybody in the computerchess world was OK to play against in principle. There was no always online service in Germany back then, so I had to log on and somehow manage it.
Bringer was kind of competive these days, and always low-rated – so there was not much of an incentive to play against it.
I am a maths guy, so I always enjoyed reading the „formulas“ of the engines on the then server. NO patch in the world has ever managed to improve +any+ engine more than the formulas on the server. I knew back then that this was not directed against me, and it sure wasn’t, but it was also funny, as it certainly influenced the opinion of the people on what engine was doing well right now.
This was ICC and the amateur world, and a time where you had to wonder at what exact time limit you may be able to challenge mighty „Ferret“,

( and this was difficult).
What you are doing now is clearly way worse. You are applying little patches and then you do the formulas with their huge influence and your „testing“, and then you make your judgments.
You are clearly wasting time and electricity .