Has anyone done a true comparison?
The kind of comparisons we have had to date are subjective. That is, a person tries both cleaning in the sink, and then tries an RCM of one type of another, and states what their opinions and conclusion happen to be.
Of those who have stated a preference after trying both, I have to say I think most have stated that they prefer the RCM, either from the point of view of convenience or the point of view of sonic superiority.
I believe a fewer number have stated the opposite, and an even smaller group have stated that they find no difference in terms of sound quality.
However, although people are entitled to their opinions, and they might be right, it's not a 'true comparison' as you ask. It's a subjective statement of opinion. Quite acceptable for discussion purposes, quite valid for the person who has tested the options and holds the opinion, but unfortunately, it proves nothing one way or the other.
There are many reasons that such opinions don't, pardon the expression, hold water.
The first is that one does not know to what extent the tester is or is not biased towards one or the other solution. They may state they have no preconceived notions going in, and they may believe that, but that doesn't mean they do not have a bias they might not even be aware of. This is a constant concern in any kind of scientific testing, and it's why double-blind testing is the rule rather than the exception.
The second is that one does not know what exact protocol is being use by the person doing the testing.
Third, one does not know if one record being tested by, say, sink cleaning, is identical to the one being cleaned by an RCM. Even when attempts are made to level the playing field in a very rigorous and legitimate way, one would have a difficult time finding two 'identical' used records to clean. It would impossible, as far as I know, to clean half the record with one method and half with the other, given the nature of cleaning in the sink (i.e., both sides get wet).
I could propose a set of protocols and a testing regimen that would resolve most or all of these issues (any scientist here could do the same), but it would involve time, labor, and effort, not to mention involving the use of expensive equipment (for measuring actual results) and unbiased third-party testers.
I doubt many would go for it. And for that matter, given the notable refusal of some in the audio hobby to believe the results of such testing anyway if it doesn't favor their preference, I doubt there would be much point to it. In other words, it doesn't matter what unbiased scientific testing would reveal; those with an emotional (or financial) interest in the outcome would refuse to accept any answer that they didn't care for.
We are audiophiles. We are passionate. We are believers. And we don't like anything science has to say about our hobby unless we approve of the finding in a very personal way. Anything else is rejected out of hand. In other words, we're about the magic; science offends us.
If one wanted to do one's own testing for one's own edification, I can think of a couple of valid ways to do a relatively unbiased test.
To remove the problem of using different records with different beginning dirt levels, one could clean a record using method A, then play and digitize it. Clean it again with method B, then play and digitize it again. Note rigorously any difference. Then take another record, but use B first, then A. Not differences.
Even though records will be different, by doing this enough times, a trend should emerge, or the results will show a 50/50 distribution. In other words, if method A is sink cleaning, and method B is RCM, and A->B shows an improvement in pop and click removal over time, then it could be legitimately presumed that your RCM regimen is more effective than your sink-cleaning regimen. If B first and then A shows more improvement in A, then the opposite. A random distribution centered around 50% would show no practical difference.
The assumption going in would be that if B is more effective than A, then the records cleaned by first A and then B would show a consistent improvement.
I won't make any assumptions about what one's results might be. I've always been willing to concede that RCMs may well do a better job at cleaning records than cleaning by hand in the sink. And if that turned out to be true, I would not be surprised, and I would not be afraid to admit it.
However, for me, the issue is not which is better. The issue is whether or not the results from sink-cleaning are sufficient for my needs. My method is indeed effective enough to meet my requirements, so I'm done.
In other words, yes, there are better SUVs than mine, but mine goes well enough in the snow that I do not look for a better one, even though I presume better SUVs exist. Make sense?