Simply put, a lot of comp players are creatures of habbit. They like their 2 Sol, 2 Scout, 1 Dem, 1 Med setup and consider switching 1 scout out for a "special case" class a tactical decision. I use this example because most "serious" comp teams are in 6v6 leagues. And the most effective setup is that.
I skimmed it, and didn't really care to read every word of all 18 pages. This over-analytical approach to the game is what drives these players. They have something figured out, that works, and all it is is a matter of "skill" at that point, at least, that's what they think and hold over us "pubstars". Which is why you hardly see a comp player "wasting their time" in the sheer chaos of a pub game.
On mapping, the fact will always remain that any system involving human interaction (and thus, tied to a complex system) will always be a complex system, and therefore, unpredictable.
Aside from the basic balancing and aesthetic appeal, a map either turns out to be fun, or not, with very little anyone can "research" to directly effect that outcome.
And for reference, maps that have been catered to these kinds of "variables", or more specifically, designed for competitive use, have all failed to be popular. (See junction, freight). Not by lack of effort in design or quality in the map itself, it just fails somehow to be "fun". Probably because the mapper wasn't thinking about how a gaggle of uncoordinated noobs could succeed/fail, but how an organized "pro" team would do it.