It isn't as easy to spot the "fakers" as you might imagine. Especially if you don't speak the same "language".
A personal example come to mind when I say this. A very good friend of mine has a very deep background in math and statistics, but from an economics background. My math and statistics background is mostly physics and control theory based. Over time we've worked with quite a few people and although I was quite able to tell people that were faking through the math when we talked an engineering language (e.g., ergodic process IIR filtering, numerical stability, etc.), but when we started talking about that same stuff from an econometric point of view (e.g., ARMAX modeling), constant translation between the two in my head made it much more difficult for me to tease out the subtle clues that trigger my BS meter...
Earlier in my dealings with my friend, I found my BS meter triggering all the time with his econometric spin on statistical modeling, but as I got to know him better, I realized my BS meter was just faulty. He knew the statistics stuff as well or better than me, but he was speaking a different language to describe the same mathematical concepts, and the papers he read had different set of seminal authors and the common data regularization procedures went by different names. I eventually took the time to learn his econometric vocabulary, but I can say I doubt I will ever be fluent in his way of talking about statistical mathematics. Having experienced this constant translation issue over time, I can say it really makes it hard to have an effective BS meter because you are constantly questioning if your own translation is accurate enough...
FWIW, I'm pretty sure have met a some real 1%-ers in my time at Caltech, and yes some of them are so out of my league that they could been BS-ing me and I still wouldn't know it. One of my classmates would sometimes look at our homework and then come up with some proof that applied some far out algebraic principle and later grin and say, well just kidding, I made that up I don't know if that proof is true, but doesn't it sounds right. Who knows if he was BS-ing or not, as he could talk circles around us in Algebraic-Category theory (and he also managed to learn how to juggle 20+ balls and ride around on a unicycle which was also way beyond me too). However, we never let him divide up the restaurant bills though as we never did fully trust his arithmetic abilities when it came to actual money ;^)