When you change pools, the luck, net space, etc. changes, so really it is not that easy to check it. xchcentral had dedicated drives (i think) for each pool (not sure how big though), and they were trying to use API to get the clean data, but we saw rather big errors in what was reported. I also had before (long time ago) separate drives dedicated to three pools, but after a month basically gave up (if I remember, at that time Flex was beating Space pool). I think few other people were doing the same thing.
I was thinking that maybe if several people would dedicate two/three drives to different pools, and we would average those data somehow, maybe that could be one way of testing it (although, each of us has a node that has different capabilities, ISP that potentially is flaky from time to time in a different way), so there are problems with that as well, unless there is a bigger sample of people.
It is kind of sad, that Chia doesn’t provide any monitoring like that in UI, so we could share that data.
Not really. I really liked Flex, but … I think that I got really pissed off during that long / low intensity dust storm, as they reported their luck almost always at 110%, and there were clear patterns that the problem was on their side, yet they never bothered to look into that. And that last issue, where they server basically stopped processing any blocks, where they claimed that it was all good was the final nail.
I think that Space is a good choice, but for me the fact that they are so big implies that they are one of the reasons that smaller pools are collapsing. Once there is no competition, they will be able to do whatever they want.
I would wait for that xchcentral to come back online (if they will come back at all), as they monitored about 10 pools, and whatever problems they had, that is potentially the best source of pools performance.