This is interesting. Thank you for bringing this up, and bring up a source that I could link to.
Since you are more familiar with it, can you elaborate on the on-target rate vs. completion rate and explain the meaning between the differential? For example, CK's 20+ yard differential is ~12% where as SL's and QE's are about 18% differential. Is there a metric which shows how QE was performing this year pre-oblique injury and post injury return, and one from last year and the year before? I think one of the scouting reports I read stated QE had trouble with intermediate passes to his right side, and I was curious if that was present pre-injury data.
Source Link.
Having read through this site a little further though, I am a little disappointed they do not include their P-values and t-statistics and residual graphs which could be used to validate the null-hypothesis, independent variable, etc. Also on the first table they present on that page, is pretty indicative that the model for intermediate/longer passes are a very poor (again though, not enough information is included here as well).
Statistics and regression analysis without the qualifiers is something I would be very skeptical about. I have seen others run regression analysis and not be able to rule out the null-hypothesis just run with it.