After reading Bill C's post on the value of returning starters, I was curious to see what affect losing or returning an All-American quarterback would have on a team's ranking. I used consensus All-Americans (from Wikipedia - some years have two consensus) and AP Top 25 Rankings for comparison. If teams were ranked out of the top 25 the next year, I treated that as ranked 26.
The results? Over the past 10 seasons, 8 consensus all-american QBs have not returned, and their teams dropped 13 spots in the AP poll on average (likely worse if you consider non-ranked teams likely to actually be lower than 26).
Over the same time span, 5 consensus all-american QBs have returned the next season, and their teams still dropped 6.6 spots on average.
My take from this quick-&-dirty comparison is that losing an all-american QB is somewhat of a big deal, and it would make sense to include something along those lines in preseason ratings (such as weighting returning starters by position and by achievements such as all-american or all-conf). Returning an all-american QB however? Not as predictive. Perhaps because it's harder to maintain the high ranking the team achieved that year regardless of who's playing QB the next. Every team with an all-american QB ranked in the top 15 that year and most in the top 5, so there's not much room for improvement.