Originally Posted by livedarklions

I've always wondered, how solid is that 25% efficiency figure? It just doesn't seem logical to me to assume that doesn't vary a lot, so is there real evidence that it doesn't?

.