On the other hand, a “proof with words” would likely have derived the explicit formulas for I_n and C_n in order to prove the relationship and would have been much more clanky.
If I had to make a comparison, this felt like reading good functional code. Concise, a bit mysterious at first, but once you understand what it does you feel enlightened. (Vs eg. wading through for loops and index arithmetic, which also does the job but is more tedious to read)
There are lots of geometrical proofs for e.g. the pythagorean theorem that need no words. It's not like code golf any more than a text-based proof is.
Words still need to get interpreted just as images do.
If anything, a succinct geometric proof can be much easier to grasp and verify than a text-based proof.
This is truly, deeply the wrong lesson to take away.
I'd say that, in a spirit similar to the Feynman technique, you don't really understand proofs whose central idea you can't explain simply -- and a wordless proof is good evidence that you have in fact grasped the proof's essence.
Of course, this doesn't mean you can do away with formal rigor. Mathematics needs that too. But I'd argue that is "the clever game for insiders to play" (albeit an important game), while intuitive understanding is what you should really be after, and is what generates new ideas and new proofs.
More beautiful proofs without words:
https://math.stackexchange.com/questions/733754/visually-stu...
https://math.stackexchange.com/questions/1337750/easy-math-p...
https://www.amazon.com/s/ref=nb_sb_noss_2?url=search-alias%3...
[1] in the fuzzy sense, not the mathematical field