Which Groupset?

Everybody seems to think more gears are only good for more range, and is apparently happy with the incredibly constrained choices being offered these days.

And sure, as you add cogs, you don't need to offer as many cassette combinations, but nobody's seemed to notice the options disappearing faster than the cogs arriving.

All this 10t stuff is madness. 11t was stupid for two reasons, and 10t is 9% worse for both of those reasons. Chordal action (the effect of the cog becoming more like a polygon than a circle) is inefficient, and it was already a thing on 11t.

The other issue is that once you're down to 1t gaps between cogs, those gaps are actually increasing as percentage jumps as you go faster, which is exactly back to front. It's a physical limitation of the system, and it sucks. The only ways to address it are to increase tooth count of cogs and rings, or to add chainrings. But of course the industry has moved in the opposite direction for... reasons?

Here's me a few years ago, thinking front shifting is getting really good, and wondering how much of a game-changer a Di2 triple would be, and oh, by the way, if you plot gear ratios, you see the line curving up once you're on the 1t gaps, which is dumb and bad even though nobody cares, and if you bust out a 38/50/52 half-step, you can make that line straight at least, which is halfway towards the ideal, which is the line curving down on the ratio graph. Incidentally, the spreadsheet I made told me that ring combo is the one - nothing else made the line as straight.

I even bought a 38t cog with 110/130 holes and some long chainring bolts and futzed around with chainring spacers to prove you can fit this 38/50/52 caper on a standard double crank, with the rings only inboard and outboard of the normal positions by 2-3mm. The two big rings can sit really close, and the 38t can sit closer to the 50 than a 39 to a 53.

I was so keen, I even programmed an Arduino to control a couple of servos which were going to operate mechanical derailers via 1mm Kevlar string, I had it mostly worked out. But of course I ran out of resources when it came to a properly-shaped FD cage, and a proper chainring setup with rigidity and shift gates...

Anyway, seems to me this is an idea which is all but destined to come to pass, since a more efficient set of gear ratios is almost certainly the biggest bit of fruit remaining on the efficiency tree. Without a doubt it's easily worth more than say, ceramic bearings or OSPWs.

I know most folks aren't quite as into close ratios as me, but someone tell me I'm not crazy here. Cause I'm pretty sure this all makes sense; it's just everyone else who's stupid, isn't it.