By all means I wasn't trying to sound too offensive, although I have been through this a lot in the last couple of days with helpers in my own scripting channels providing "crappy" code.
And yes I was referring to the mass use of tokenize, since it can be much, much slower. In this case, tests returned that the tokenize'd alias is twice as slow as the other variant (31ms and 15ms respectively), that is why I was inferring that it is indeed slower.
Although in this case, as you pointed out, it doesn't make that much of a difference (only 16ms), I was just pointing out that that style of looping is a very bad coding habit.
/tokenize should never be used in looping where another way exists.
