I believe that is what I said, 20 years or so of mediocrity and little or no national relevance.
If Michigan went 20 years without ever being ranked in the top ten even midseason, I think they would be dropped. I think if they go another 10 years with 9-4 and 8-5 kinds of seasons interspersed with a few non-bowl years, they get dropped, by most.
If they insert some 11 win years in there, they remain on the radar. It takes sustained mediocrity perhaps more than really bad years and really good years. You can't be an 8-5 kind of program and remain a helmet forever.
I think you will agree with this (and you somewhat incorporated it with your "20 years without ever being ranked in the top ten even midseason" comment) so this is not an argument but a clarification.
I think the key issue in determining a given program's continuing helmet status is the issue of relevance. Therefore, I am hesitant to say, as you did, that you "can't be an 8-5 kind of program and remain a helmet forever."
My first reservation with that statement is that it depends somewhat on SoS. If you go 8-5 with two OOC regular season losses to top-10 teams and a bowl loss to a top-10 team and only two conference losses (which are likely to both be top-10 level losses) then you might be a darn good and quite relevant team. OTOH, if you only play two top-10 level teams, lose to both of those, and also go 8-3 against your 11 unranked and barely ranked opponents then you aren't very good and you aren't relevant.
The second reservation is timing. You said 9-4 and 8-5 seasons for the next 10 years wouldn't be enough for Michigan to drop. My response is that it depends. Look at their 2017 schedule for a great example:
If they go 9-4 or 8-5 by losing to Florida (first game), Penn State (mid-season), and Wisconsin and Ohio State (last two regular season games) then I agree, that isn't maintaining their helmet because they'll basically be irrelevant all year.
- The Florida loss will knock them back to (unranked/barely ranked).
- They'll head into PSU at 5-1 and closing in on the top-10 but losing will knock them back to (unranked/barely ranked).
- They'll head into Wisconsin at 8-2 and closing in on the top-10 but losing will knock them back to ~20.
- They'll head into the Ohio State game at 8-3 and probably ranked but losing will knock them back to (unranked/barely ranked).
Now consider instead the alternative that they go 9-4 by splitting the UF/PSU games and losing their last three (UW, tOSU, bowl):
- Coming out of the PSU game they'll be 6-1 and top-10.
- They'll head into Wisconsin at 9-1 and probably top-5 or close to it. The loss will probably leave them either barely in or barely out of the top-10.
- They'll head into the Ohio State game at 9-2 and ranked around #10. The loss will knock them back to a lower ranking.
- They'll head into their bowl at 9-3 and still ranked. The loss will leave them (unranked/barely ranked).
In both examples the Wolverines finish 9-4 but how they get there makes a big difference in terms of helmet. In the former example the Wolverines are basically irrelevant all year and their helmet diminishes a bit. In the latter example they are highly relevant for most of the season and their helmet is reaffirmed.