It will also be the slowest only 1/3 of the time, so in fact this isn't an explanation why "the other lines" seem to be faster.
I think the effect is the result of fixating on the fact that there was a faster line, not that there are slower lines.
I'm interested whether that would mean that Monty Hall problem solution can be applied here and whether switching lines would actually improve your chances for being in the fastest line.