So far, this seems to say something about the voters on that old poll: they weren't so dumb.
Note the near one-for-one correspondence between the order of choice popularity and the order of returns so far.

Interesting observation. Intriguingly good fit.

But perhaps, for a poll like this, returns at a certain point in time also has something to say about volatility?
Cannot help but notice the giant shooting star on SNOW on December 8, 2020, as well as the rapid run up on TSLA.

So, out of curiosity, I looked at the update on this poll in two alternate universes.
In those, the poll was posted on a Thursday a few weeks earlier. 12/11 instead of 10/12.

Returns as of Friday 11/11/2022:

        at open  at close
BRK.A   +36.8%   +39.5%
SNOW    -39.1%   -32.2%
TSLA    +34.4%   +42.8%

(Pen and paper math, so please forgive me for not including all of them.)

Which makes me wonder. Could there be a less noisy, yet fair and agreeable, complementary way to calculate interim results for a long-running challenge?
Something simple that tries to take random spikes and volatility into account, while still being true to the original poll?

Nothing important.. just, well, there's a dilemma in this that I find interesting.