Welcome to Syntactic Structures, the Standard Theory (tm) of how linguistics blogging should be done. Your host is the leading authority in whatever field of linguistics he's decided to pronounce on today, and whatever he says -- goes. In my spare time, I allow other people to maintain a blog full of thoughts I agree with on non-linguistic matters.
We've got a lot of ground to cover here, folks, but we'll start with some good old-fashioned score settling. The first thing you need to know is that this other blog is basically wrong about everything. For example, there are no such things as snowclones, which are another example of what some people call the "lexicalist hypothesis", a view that has occasionally -- but erroneously -- been inferred from some previous writings of mine. Since my linguistic creativity is infinite, I can categorically state that there aren't templates of any sort.
I have, however, recently discovered a phenomenon to which I have given the creative and original name of "hailimitations". These are cases where people coincidentally come up with the same pattern, with just a few substantive nouns changed. The name derives from the fact that the first time I happened to come across this pattern, it was in a news item from the Economist, and said "If Eskimos have dozens of words for snow, Germans have as many for bureaucracy." By coincidence, other people have also used a similar string, as in this piece which I independently uncovered, "If Eskimos have N words for snow, then Santa Cruzans must have even more for surf." Since the sentence are almost -- but not exactly -- alike, we might say that they are imitations of each other. That is, if each one was not carefully and lovingly fashioned through the perfectly autonomous innate gifts of syntax unique to every human being. Admittedly, hailimitations would be on far more solid ground if I had come up these examples through introspection, and not by using so-called "empirical" data, but c'est la vie.
Hailimitations are not, however, to be confused with the wholly distinct theoretical object which I call HAILIMITATIONS, which are sentences whose I-language representation includes the concept "hail".