Comment Re:what I did (Score 1) 510
Oh well.
My two examples were really of one principle: Code's look should be a hint to its function, and violating that is a problem. The point of the first example, for instance, was not that the 'z' line is necessarily incorrect; the point was that the programmer who wrote it was unclear about her intentions, effectively saying one thing to the computer and another to human readers. This being a problem is not a myth; it's common sense.
Your claim comes down to: No, such violations do not matter (or never happen in practice). You even claim to have some quantitative argument to support that. I doubt the argument will hold -- most such studies conflate to many factors to make strong positions on one language feature or other. And barring such data, your position seems quite irrational.