Screen readers can surely handle Greek just fine, the only problem with the given spelling is that it uses "lookalike" glyphs instead of corresponding characters from the Greek alphabet (eg. use of "mu"/µ for "u").

Of course, I am sure some screen readers will simply take a custom text preprocessor that will turn l337sp34k and similar spellings like this one into regular text, but I am certain none of them have it as the default.