Interesting, although still inconclusive. The big problem is aiming. The article's approach is, more or less, to aim each mirror separately, then cover it up while aiming the others. Hence it took quite a long time, which is problematic when the target is moving and you are in a battle. Maybe it just needed more investment in men and materials - which I believe Archimedes would have had available. (One problem with Mythbusters is that they only spend a few days on each Myth, and have limited resources too.)
Some questions: what's the most number of mirrors you can aim at once, before you become unable to tell which bright spot is due to your mirror? How long did each mirror take to aim? It seems like you could do maybe 5 at a time, and with practice aim in maybe 5 seconds, so 100 mirrors would take 100 seconds, which is perhaps fast enough to be useful.
Maybe Archimedes experimented with the technology, decided it was less practical than shooting fire arrows, and leaked the experimental results to scare the enemy. Kind of what the US did with some of the "star wars" experiments.
_________________________
Quality is addictive.