Wordcount isn't choked by mechanics that serve a purpose to do a thing you don't like. All of those Charms have a purpose and do a thing. That is like me accusing Scion of wasting wordcount on Paths because I feel it restricts flexibility in character design. No, it just does a thing I don't like.
The reason it chokes the wordcount is because it's largely redundant. They
don't do new and different things. They are often rehashes of similar, if not identical, dice tricks in other Charm trees. That's the thing that is a step backwards from 2e to 1e in 3e: they fell too much in love with the notion of manipulating the underlying mechanics, and lost sight of it actually being a meaningful use of word-count.
I won't deny that you might lose some tiny nuances between Charms if you simplified it to a shorter list of dice manipulations and consolidated that into Excellencies, but I don't think the slight gains are worth the massive wordcount and thus enormous inflation of page count and cost-of-printing that it costs to gain them. The mechanics between them just aren't unique and different enough to really warrant it.
It's not that they do something I don't like. It's that they spend so many words on so little effect. And worse, that all those words get in the way of finding more unique and interesting effects, because you can't easily parse the enormous Charm trees to differentiate "yet another die manipulation" from "something new and different." Worse, you can't look at the Charm names and determine whether this is "die manipulation trick aleph" or "die manipulation trick omicron." Or "like die manipulation trick bet, but sliiiiightly different in that you can't keep as many of the altered dice."
While those do tell a story, it doesn't stop other dice tricks from telling other stories
Of course not. But there comes a point of diminishing returns on the actual quantity of different stories to be told and the number of Charms spent making up dice mechanics to tell them. 3e has far passed that point.
2E's structure was
sufficient to tell the stories needed. And nothing says you can't have a few more dice manipulation tricks, if you absolutely, positively feel this particular one is new, interesting, relevant, and does something that is needed by the splat. My point is that 3e stepped away from the coherent, neatly packaged and well-organized method of unifying all the "same dice trick, done again" Charms, in favor of, well, "same dice trick, done again but
slightly differently" Charms repeated over and over and over. If my eyes didn't glaze over with "wait, what does this actually...mean?" after the umpteenth dice trick Charm in a tree, I'd have less problem with it. And I
like fiddly mechanics, as a general rule.
There is interesting stuff in the Craft rules they've devised, for instance, even though I think they're sadly not as playable or useful to actual gaming as I would like.
The fact that I can't read the Charms in the 3e book without getting lost and bored and feeling like I've read all this before earlier in the book is the problem. It is all too clear that the designers were fans who got caught up in their "vision" or their "hey, this is cool mechanics" stuff, and couldn't bring themselves to pare down and edit for coherence and conciseness.
This is unfortunate, because they had some really good ideas. They just needed to have somebody who could point out the flaws and help them consolidate, rather than letting them spin out and out and out without actually adding more to it.