> I went in with doubt I had done it right.

And if in your doubt you decided to run it through the interpreter to get the "real" answer, whoops, you're rejected.

That's cheating (even if it just assures you that your answer is correct)

Is it? The page implies it's allowed, but they want people who think running it is "more of a hassle".

Oh right, it seems to be allowed.

I don't know then. I can open up a terminal with a python and paste it really fast, faster than run it in my head.

That doubt is valid. Anyone reading this blog post (or in an interview, given the prevalence of trick interview questions) would know there must be some kind of trick. So, after getting the answer without finding a trick, it would be totally reasonable to conclude you must have missed something. In this case, it turns out the trick was something that was INTENDED for you to miss if you solved the problem in your head. At the end of the day, the knowledge that "I may have missed something" is just part of day to day life as an engineer. You have to make your best effort and not get paralyzed by the possibility of failure. If you did it right, had a gut feeling that something was amiss, but submitted the right answer without too much hemming and hawing, I expect that's typical for a qualified engineer.