For a large branching game, that’d be a huge job. You’d have to go through the game heaps of times, tracking all the different possible combinations of choices and decide where to draw the line between an “important” choice and one that’s providing flavour text or a stat change without extensive text. There’s a bit of a grey area there, I’d argue that having at least paragraph or two of text makes it a valid choice in many cases even if there’s no stat changes.
I think it is possible to get an idea of numbers of choices per playthrough using the random test and hand counting them for each seed. It won’t tell you much about the quality of the choice content though 
Edit:
Nope, I remember Jason saying that it was only 23k. (It’ll probably turn up on a forum search from memory.) Some games are deceptive like that for their length and seem longer or shorter than they are
Edit2: Here it is
(How times have changed over a few years, ~100,000 words used to be considered a decent length for a choice game (The original Heroes rise is 110,000) 
)