(Gonna try to do the best I can here using only info that's been made public before)
"While at Google, only one person cut my interview short. I thought it was weird because I had answered her question correctly. But apparently they often write off a candidate for not using the right terminology or some other picky constraint."
I see the other side of the interviews (i sit on a large number of Google hiring committees, and have for about 7 years), and I certainly don't see this happen often. Mostly among newer interviewers. HC gives feedback when we see it, and that puts a stop to it. We either ignore or adjust our thoughts on the interview score to account for it.
Without knowing your exact case, i'll point out there is rarely a "correct" answer to the interview question. It's usually more about thought process and whether you can reason about how to improve upon solutions, etc. Hiring committees hate interview questions that rely on tricks or whatever in order to get good solution, and usually either ban the questions or at the very least, ask interviewers to change questions (and again, adjust our thinking to ignore or whatever the interview score)[1]
Hiring Committees certainly try to make sense of interviews, and we are, after all, people. It's not like we look at interview scores, go "welp, they are 0.1 points below the average hiring score, so that's a no". We are also very good about spotting inconsistencies between interviews, etc.
I'm not going to claim it's perfect, but it's definitely not what reddit claims, where
"They look at it for 3-5 minutes and decide whether you are hired or not. In exceptional cases, they decided that they don’t have enough information, and ask you to do more interviews. "
I spend a large number of hours a week at night reading candidate interview packets, writing up what i thought of each interview and how it affected my thoughts on hiring the candidate, etc. I then spend an hour in each committee discussing candidates.
It may be that in committee, it takes 3-5 minutes to make a decision, but that is because committees tend to end up with very consistent decisions between members (and we don't see each others initial decision or reasoning until the meeting starts). If the decision from everyone was "no, we shouldn't hire", or "yes, we should hire", then there is nothing really to discuss.
If there is dissent one way or the other, long discussions over candidates often break out, and we keep going till we come to a consensus about hiring or not hiring.
Usually we don't need to ask for more interviews, there are rare cases where we just don't have enough signal in the interviews.
[1]
I believe it was Rob Pike who said that it is very easy to come up with questions where if the person does badly, you don't want to hire them, but very hard to come up with questions where if they do well, you do want to hire them. Questions that have "tricks" in them basically never fall into the latter category.