I don't know the particulars of the situation of course - I imagine there's a big bump in work involved between making a tool useful for its creator and making it idiot-proof for one - but I have observed that as organisations get larger predictable performance becomes much more valuable relative to outstanding performance.