If there was some sort of advantage to having a $9M player sitting in the press box, and secondary assists were just as good as goals, you'd totally have a point.
But here in real life they've been in the NHL for the same number of seasons, and goals are far more important than secondary assists.
Since 07-08 (sorry, nerd stats only go back to 07-08), OV has a RelTM/60 of 0.40 (total of 116.06) at 5v5 and of 0.32 (total of 27.47) on the PP. Crosby is at 0.62 (total of 155.01) and 0.62 (total of 41.19) respectively (funnily enough).
I generally agree with goals > primary assists >> secondary assists, up to a certain extent. We're not comparing Jason Robertson and David Krejci, but two of the greatest in history at their craft. Crosby, despite his lower goal total, is a much greater overall play driver, and overall his assists cannot be deemed, on average, to be as replaceable as the "average" assist. The RelTM stats I referred to, while in no way perfect, do have a meaning due to the sheer sample size - and show that Crosby does seem to increase his team's / teammates' on-ice overall scoring (so his goal + his other teammates' goals) more than OV did. Obviously, the Caps have played a more offensive style overall, and it may be harder for a star player to separate from their teammates when the team as a whole is offense first, thereby potentially OV's numbers there.
Looking at some more advanced stats, EH's model puts their "career" (i.e. since 07-08, so not their entire career) Offense GAR (so just the offense contribution) at 222 for OV and 227 for Crosby. Their Offense xGAR (again, EH's model, which I tend to prefer to their GAR model) puts OV at 283 and Crosby at 272, with OV having the edge on the PP and Crosby having the edge at 5v5. Even taking into account injuries (otherwise as you know, Crosby would have the edge, but I agree "availability" is a thing), this is, like, very, very close.
Unfortunately, I can't bring up those stats for their first 2 seasons, since they simple aren't available.