I guess I don't really see your point. Why are you so certain that the LVSC computer model mirrors the "public predictors"?
Your whole argument seems to be that you are certain that the line was manually adjusted by some guy with gigantic kahunas that is likely better at figuring spreads in his head than a wagering driven computer formula that's been proven over literally thousands of games.
amirite?

Because you're linking two independent wagers together and saying you can devise what the computer does with simple math. The computer sets the spread with the goal of betting parity. It sets the total with the goal of betting parity. The computer has no need to predict a spread or a total.
And my argument isn't that the line was arbitrarily set by one guy, it's that the computer, guy or chicken picking games isn't concerned with predicting a final score to provide the spread and total so trying to use "simple math" is foooocking Fake Sugar Dick (WARNING, NOT THE REAL SUGAR DICK!).
The point is NOT whether or not the books are trying to "predict a final score". They're not. It's pretty common knowledge that the goal is to limit overall exposure.
However, for all practical purposes, one can easily deduct an approximation of what they model sees as a final score if everything goes exactly as the model forecasts. Although separate, it's foolish to think that these wagers are totally "independant" as BOTH are on the SAME contest.
Therefore, an opening total of 52 with an opening line of 10 simply implies that the model used would be approximating the score at about 31-21. My point is not that they are trying to predict the score, but that one can easily come up with that approximate score from the two factors provided from said model.
It's even more rough ridin' Fake Sugar Dick (WARNING, NOT THE REAL SUGAR DICK!) not to recognize and realize this inherent relationship in the two numbers.