You're dead right on this, broadly speaking. The appeal of viewers isn't so much by market size but market demographics, age, income, etc, but that all viewership isn't created equal is true.
The thing you're missing is that the actual eyeballs are much less regionalized than you think.
A lesser Florida State game isn't going to outdraw a better Clemson game in Miami. It might not in Orlando or Tampa either. The game that wins the week wins nationally.
When forcing grannies who never watched sports in their lives to pay $1 per month in perpetuity for the BTN you go from zero to 1, it's all-or-nothing for the entirety of a TV market.
When we're talking about actual eyeballs it's much more shades of gray, to the extent there is even regionalization which there largely isn't.
So the raw viewership total uber alles isn't *exactly* the end of the story, but it's a heuristic that works dramatically more reliably than "ah, if we show Oregon games that will mean we now have Pacific Northwest viewers"
Good points, and I think you are correct. However, let's say you could broadly divide the viewers for any given game into three categories that I am totally making up now:
1. The diehards for each team that will tune in no matter how good either team is or what channel the game is on.
2. Casual fans "predisposed" to hop on the bandwagon for a given team but the kinds of fans who won't tune in if the team is bad and/or are more likely to watch on ABC or FOX simply because watching those channels is a habit or because they're in a market without channels like BTN or ACCN (e.g., many of the people I have met in Chicago who only watch Illinois when they are good at something but then sort of "claim" the Illini as a home team of sorts).
3. Casual viewers in general who just tune into big college football games because they enjoy the sport and prefer to watch "good football."
Group 3 is tuning in (in somewhat fluctuating numbers, I will admit) to most "big games" ... they didn't care about Clemson in 1999, but now they think watching Clemson is exciting. Group 1 fluctuates
somewhat based on things like size of your alumni base and the local population, but I think it is mostly the truly massive fan bases at the top (Notre Dame, Texas or OSU), the seriously small fan bases at the bottom (Northwestern, Georgia Tech or Vanderbilt) and the vast majority of us in the middle.
So, that leaves Group 2 ... and I really do think that might be a separating factor for many schools as far as "potential for drawing great tv ratings" goes. In a hypothetical comparison, let's say Illinois and Oklahoma State are relatively even in these categories, with Oklahoma State currently having a larger group for Group 1 and the two schools not being materially different in how they draw viewers in Group 3 if it is a relevant matchup featuring one of our teams. I think what might make Illinois more appealing is that the casual fans we have proven (albeit in isolated instances) we can draw when we have a truly exciting and good football team are coming from a much larger pool of potential viewers than Oklahoma State can draw from ... in other words, I think it's easier for a resurgent Illinois to pile on additional viewers much more quickly than, say, a 10-win Oklahoma State team vs. whatever their record has been lately.
This is all just my conjecture, but I would assume that an "untapped fandom" population is heavily considered for schools who are below the mega-TV draws like Michigan or Alabama. I think schools like Iowa State and Kansas State have maxed out their viewers in a way that schools like Illinois or UCLA have not.