... In order to drain you have to crack the third loose, that seems absurd to me.
Yeah, to most of us too. But I think the reasoning was that to drain the lube (which back then was not considered regular maintenance by most people) this way was a good reason to pull it all apart and check the brakes, and more importantly the rear wheel bearings.
If you only did your lube every 100k miles, you just did it all at once anyway.
In reality, they were probably just trying to make it simpler, cheaper and stronger with fewer holes. But the secondary result of having to check wheel bearings and such just to change your fluid was a good result.
Or maybe they expected you to do it through the pinion support all this time anyway?
Right now I am testing my drain for leaks. A bit of water with red food coloring, paper towel below should tell me if it dripped. Correct me if I am wrong, if water does not leak, I should be good, right?
Not so sure. Water might be "thinner" than oil, but it's got different lubricity properties too. And what about surface tension? I bet some oils will find chinks easier than water in some circumstances.
But that's only a guess. Maybe it's more to do with the duration of the test. If you let water sit for a day and it doesn't leak, that might not mean that oil sitting behind it for a month won't leak.
Anyway, I'm no chemist, and oil does float on top of water (generally), but not sure how the whole leak testing would vary because of it.
It's probably as good a test as you're going to get. Hmm, or maybe try some solvent instead of water? That might be the medium that gets the quickest results.
BTW, got pics of this creation? Be great to see what you did. Lots of people have done it, but even more still wanting to do just that.
Paul