At good places you sometimes run into professors that you have difficulty understanding. At bad places they won't recruit professors that you have difficulty understanding because the darling students will complain.
That complaint of yours - it says more about the fellow who makes it that it does about the TAs.
I had a calculus TA at Berkeley whose English was so poor that she would respond to questions from Chinese students in Chinese.
When she didn't understand your question in English, she would make a blind guess as what you were asking about and then answer that, speaking as fast as she could so that you couldn't interrupt her to tell her that she wasn't answering your question. This would continue until you gave up.
I had a similar experience in an Algorithms class, where the TA had a firm grasp of perhaps 70 % of the material she was supposed to teach. This person was an American citizen, though.
When I was discussing the quality of TAs with the Instructional Director in our department this episode came up, and she was deliciously cynical. "The students are resilient", she said.
This is true. At university you are supposed to work yourself and to create learning networks with your fellow students yourself. It isn't highschool any longer, especially not at Berkeley.
The point of a university is to master the subject by any means necessary, which includes building up your professional and social network. This point was impressed upon us in our first week at uni. We were encouraged by profs and TAs to set up informal learning groups, if we didn't we weren't going to cope with the pressure. I keep telling my students the same.
I hate to say this, but it must be said: anyone who calls dealing with an incompetent TA "hazing" when he is at a genuine university like Berkeley has missed something. Didn't you have fellow students to talk to to be able to pass Calculus?
I know there are not-universities, and anyone who passes through these is cheated, but Berkeley isn't among the not-universities.
I sympathize very strongly with your viewpoint (and had my own horror stories with both domestic and foreign instructors, some tenured, some TAs, as well as some very good ones).
That said: among the most powerful and valuable lessons I learned in college, and ones I really wasn't prepared to even understand were part of the "curriculum" involved dealing with combative instructors (getting screamed at, in the locker room showers, while both of you are nude, by your professor is ... an interesting experience), or complete foulups in financial aid disbursement (calmly telling the woman who told me I'd have to call another department, which had just closed for the day, after I'd realized where the check I should have received a month before had ended up, what my demands were, and picking up their distribution the next afternoon, felt pretty good). Relationship ups and downs. Deaths in the family, of family of other students, of other students and friends.
It's not like college was a war zone (though it had elements of that), but there were definitely some learning experiences.
And a few skills I picked up, many of them outside my core coursework, or even any coursework, which I still practice daily now to great profit.
However, when the same TAs are responsible for grading and explaining assignments, it doesn't really matter whether a student is 'resilient.' No matter how much effort the expend, it's not getting reflected in actual knowledge if the TA is grading mostly on rote format because of language barriers and the students fail to grasp any significant content because it relies on a broken intermediary to be communicated. It becomes an exercise in showing up to lab, sitting there, doing the the things the TA says, putting the exact tables they specify in your report, and being oblivious to what the lab was supposed to be accomplishing. Or at least that's what I witnessed.
What you say is somewhat true for classes with regular textbooks or a prof to go back to, but when it's a lab where all content comes from the TA running the lab, often times you end up going through wrote procedure that's potentially even completely disconnected from the intended lab material because it's been intercepted by three layers of translation and language confusion.
What you are describing is a clue problem, not a language problem - the root cause is that the TA does not understand the assignment, if they can see it and still don't understand it isn't the language that's the problem. What you are describing is familiar, it is just like in that Algorithms class I sat through.
The original comment was about TAs with poor command of English.
Almost. It's more like "At Doldrum Hills State a fair amount of undergraduates are decent, but their graduate program is shit, and everyone in there is awful."
That complaint of yours - it says more about the fellow who makes it that it does about the TAs.