Of course it's semantics. Anytime you're nitpicking details of a definition you pulled out of a dictionary, you're playing semantics. Especially if you reach a conclusion that is direct contradiction to a generally accepted understanding of what "market share" actually means.
Unless, of course, you want to pretend that only dollar market share is market share, and not unit market share. I don't know anyone who knows anything at all about marketing who would agree with that, though. Unit market share is about number of units in the market, and whether or not they were distributed through a "sale" or a freebie is immaterial. As is the cost ($0, in this case.)
Plus, now you're arguing about my "and another thing!" point. The original point was that comparing discussion of a product that isn't officially released in a final form to one that is is a meaningless comparison. A product that is due to come out soon should b expected to generate chatter, but that says absolutely nothing about any putative correlation between chatter and sales of products that are already out and have been for some time. Therefore, using that to "disprove" the idea that online chatter is a proxy indicator of the use of the product in the market, is nonsensical. Imaro's claim was an attempt to be clever, but it failed at two levels because 1) it was not an apples to apples comparison, so his "proof" of the flaw in the methodology is meaningless, and 2) it's not even true anyway, because D&DN is, in fact, already present in the market--at least in a beta version. The more we dance around those two points to try and salvage his argument, the more we wander further and further into la-la-land.
Aaaaaannnndd... with that, I'm officially done with this completely stupid conversation.