The whole "recruiting rankings don't matter" is probably the most wildly inaccurate argument that gets repeated on this forum.
They are on average correct but with wild variability for individual players. A single guy who is rated 90 probably only has about a 60% chance of being better than a single guy who is rated 85. But if you repeat that over 20-25 players per year for 4 years, there is a 100% chance that a roster full of 90's is a better team than a roster full of 85's. Literally 100%.
Here are the class rankings the last 4 years for the teams that made the final 4 of the CFP:
OSU 5, 4, 4, 2
ND 9, 12, 7, 9
Texas 6, 3, 5, 15
PSU 15, 14, 6, 21
Other teams of note:
Oregon 3, 9, 13, 4
UGA 1, 2, 3, 6
If you are not bringing in top 10-15 classes every year you can absolutely forget competing for a national title. Pointing to one guy where the rankings didn't pan out is a complete misunderstanding of statistics -- it's like saying Vegas has no idea what they are doing because they got the line wrong for one game. Ok, then try betting against Vegas 80 more times