The math regarding Cryptic Command and U sources eludes me. Here, it states in order to have a 90% cast success rate in relation to U count, one needs 22 sources of U for a T4 requirement of UUU.

This multivariate hypergeometric distribution calculator, the math of such explained here, deems that 4 Cryptic Commands require 24 sources of U for a T4 Cryptic Command. Which is quite surprising, as typically, for high card counts the calculated source count for a 90% success rate is quite similar between the Frank Analysis tables and the multivariate hypergeometic distribution tables, but there's a difference of two here.

Even more confusing, I've seen competitive lists from professional esteemed players, playing anywhere between 16-22 U counts when running 4 Cryptics. Granted, the 16 U list was running 4 Aether Vials alongside Eternal Witness(softlock with cryptics), but it's apparently common for even Jeskai control to run 19 U for 4 Cryptics.

I don't understand why this is the case, and I'd like to generate a discussion on it.

Does card-draw from cards like Remand, slightly reduce at all the need for U counts? I see that drawing cards increases our likely hood of drawing a U source, but it also increases our likelihood of drawing a Cryptic, so I'm unsure about this.

I understand that, we won't have 4 lands by T4 90% of the time, anyway. Do we need less coloured sources to support the high-end part of our curve? I have been thinking the opposite, that in order to have our higher end of our curve not get clogged in our hand and be worthwhile to play, we must be able to consisently cast it on time.

Can it simply be that the lists that play 4 Cryptic Commands, gain enough value by such, and are slow enough, that they're okay with not playing it consistently on T4? That the trade off for playing more U sources, or less Cryptics, simply isn't worth it, for whatever reason.