Zaikesman, your last point, "I'd have to give a shit," is exactly the point of those suggesting that the CLC should be tried, especially given some reviews that suggest it works. We don't really give a shit that some demand an explanation, if we hear an improvement.
I have tried several tweaks that did not have very good explanations, notably the Bybee filters and various cd mats, that proved of no benefit, but I have tried others that did, such as the IC, which I heard demonstrated at CES and the Muratas also demonstrated at CES, that proved excellent. None of these have been in experiments where statistical significance was assessed. It matters not to me if I hear a difference.
I teach research methods and find the question of whether a random sample's sampling error could have accounted for the variation noted to be trivial, especially as anything will be statistically significant if the sample size is large.
I have tried several tweaks that did not have very good explanations, notably the Bybee filters and various cd mats, that proved of no benefit, but I have tried others that did, such as the IC, which I heard demonstrated at CES and the Muratas also demonstrated at CES, that proved excellent. None of these have been in experiments where statistical significance was assessed. It matters not to me if I hear a difference.
I teach research methods and find the question of whether a random sample's sampling error could have accounted for the variation noted to be trivial, especially as anything will be statistically significant if the sample size is large.