Picking one of my leagues at random... the average amount of points per WR start in that league was 18.25. WRs 22, 23, and 24 (by points per game) produced 18.4, 18.2, and 17.6 points per game. So, pretending I had another WR of exactly equal quality to cover all byes and injuries, I would say that WR22/23/24 would be pretty dang close to a league-average unit.
Now, let's say I replaced WR22 with WR12. In theory, WR12 is the absolute worst WR1.
My WR3 shouldn't be measured against a different baseline than my WR1.
If I drafted Jerricho Cotchery in the 18th round, it's not like that's a better draft pick if I also grabbed Josh Gordon in the 10th. Josh Gordon being amazing only made Josh Gordon a great draft pick, it didn't magically make Jerricho Cotchery a great draft pick, too.
I think you're taking my statements too literally. My stance was not that we need 3 different numbers, or that Dez Bryant has more value at WR3 than WR1. My issue was with using worst projected starter as baseline production. Functionally, there is nothing "baseline" about it.
ETA: I'd like to see a league in which WR22,WR23,WR24 is a baseline unit, if you don't mind sharing the link.
I'll PM you the link, but honestly, it's not as surprising as you'd think when you think about what I'm measuring. It was a 10 team league that started 3 WRs a week for 16 weeks, good for 480 WR-Starts. Those WRs scored 8763.6 points, good for 18.26 points per WR-Start. In this particular league, 23 WRs produced a higher PPG average than 18.26. Which makes total sense, because a lot of those WRs with higher PPG averages didn't play a full 16 games (Blackmon, Julio), and a lot more of them spent some time on the bench before their owners trusted them enough to start them (Gordon, Jeffery), and some teams had terrible WR corps and brought down the average (one guy's best receivers were Julian Edelman, Reggie Wayne, Victor Cruz, and Nate Washington).