Hmmm that could be interesting. Maybe I could do GP+(GS*2)+ Assigned value for pro bowl appearances. Then you could literally see the "value" of each pick and then say "oh, a first round pick around the 14th pick should have this type of career". And you see picks that overperform for their slot and are thus good picks. I might look into that, even though, as you said, it would be quite tedious.
I agree you could definitely go in a lot of directions with this depending on how and what you wanted to evaluate a guy on. I think the pro bowl and other stats would be very meaningful in evaluating 1st/2nd round guys because a lot of those guys will get playing time and there needs to be some other variable to differentiate players.. I think you would probably want to adjust your logic based on what type of comparison you are doing so long as you can get to some level of quantitative analysis.