Right now, early elem classes are 30-35. To me that is huge and class size is often used to justify lack of oral and group learning along with lack of differentiation for above and below average students who don't have an IEP and reliance on mountains of worksheets and daily homework to "measure" learning. To put it in context, when my 10th grader was in K he was in a class of 16 with a teacher and a teacher's aide. When my 8th grader was in K, it was a class of 24 with a teacher and a teacher's aide. If we'd sent my 1st grader to the public K it would have been to a class of 33 with just a teacher, the aides having been laid off over the last few years. I agree that language-based math offers an improvement for those who struggle with math but excel in language arts (and being one of those people, I certainly see the value in that). However, what about those kids who struggle with language arts? I haven't seen it offer any improvement to them. It just undermines their confidence in their math ability and by extension their overall intelligence because they're getting marked wrong for correct solutions presented in incorrect language or method. Why can't they be allowed to excel in the black-and-white world of math while receiving the support they need to catch up in their language arts classes? We don't demand that English lessons include math concepts to reinforce lessons across the curriculum; where is the sense in demanding that math lessons include English concepts? As to geometric proofs, those don't require good grammar, spelling, or punctuation. My son excels in math now that he's reached the stage of proofing formulas (he's in geometry now). It was the stage of "answer this story problem in a complete sentence and explain how you arrived at your answer" that made him want to throw in the towel.
Our class sizes are fairly comparable. The only reason they are generally between 25 and 30 rather than 30 and 35 is the physical size of the classrooms.
I don't think the arguments you are presenting about this making kids who are good at math want to throw in the towel sound any different to the ones I saw presented years ago, and the ones I actually tried on the teachers myself, about showing the work behind math problems, even the very simple ones. I hated having to show my work, but honestly as the math got harder being forced to do it all along was of benefit. Your son may have been very frustrated in the past but doing that work might be a part of why he is excelling in geometry now. As to whether it undermines the kids who are good at math but poor at language, I would like to see some evidence of that. The exact effect on a single child or group of children in a single class, school or district isn't what I'm talking about here, simply because those effects are at least as likely to be down to specific teaching methods or the chosen curriculum. I don't see that the standards themselves do undermine the math-loving kids while providing them no benefit.
Language is more than just English classes. It is how every subject, math included, is communicated. We have created artificial distinctions between subjects that are not nearly as clearly distinct as we act. Lose out on math and science will become increasingly difficult, particularly physics and chemistry. Being unable to write effectively will hurt in all of the other subjects. Except in rare situations, a child who can learn to write out the equations for math could also learn to write it in English sentences and the insistence that doing so "isn't math" is likely the biggest barrier. A child who cannot at least verbally explain the method used likely does not have a clear understanding of the generalised method but only the specific case of that problem. A child who can add 1 to a number but cannot yet add 2 does not understand addition generally, but only the specific case of adding 1.
Funnily enough, my son had math homework this week (they cover 2 topics in homework each week, more in depth than the equivalent I did growing up). He had to show what he had learned from his mistakes in a math test they took in class last week. He chose to cover perimeter and area of rectilinear quadrilaterals (yes, he honestly used those words!). He did so both in concrete examples of a square and a rectangle he drew on graph paper then worked out the perimeter and area for each of them. He also wrote, in sentences, the rule for finding the perimeter of both a square and a rectangle and the rule for finding the area of each. Because he has always been expected to explain his math, he didn't consider that difficult. If it matters, his other topic was researching some of the Egyptian gods and goddesses.