The syllable count fails on some words. For example "naïve". If you can't trust the program to count syllables at least as reliably as the writer using it, what use is it? Interesting idea though.
Ironically, the diaeresis which it probably ignored is essentially an indication that the two adjacent vowels do not form a digraph and hence form parts of two syllables. (Admittedly, umlauts and diaereses are represented by the same characters on a computer, but their placement with respect to other vowels gives them away most of the time.)