The tape is critical. However, folks play in different systems and against different levels of competition. Lots can change after the end of the season.
1) post season games and practices
2) health level (obviously critical for those who could not play a full 2009 season)
3) combine numbers (how fast, quick and strong are they really) and pro date numbers (another chance to show that one really isn't that slow or stiff).
4) individual tryouts (an opportunity to test skills particularly important for your team)
5) background checks and individual interviews (these are critical in assessing off-field issues, character issues, and attitude issues.
An example of the value of the offseason is in the evaluation fo wide receivers. Just based on tape, the conclusion would be that some players just can't haven't run a full route tree.
In the several offseason opportunities, receivers have an opportunity to demonstrate what they can and cannot do, and what learning still needs to take place. Some wide receivers are pro-ready, some are not. Some can start as a slot receiver, some (most?) need a year or more of seasoning and nfl experience before they can be counted on.
===================
All the above being said, many analysts and many teams underestimate the importance of actual production on the field, and overestimate combine and other numbers. This has gone on for a very long time. There have been many workout warriers and track stars that have failed in the nfl.
I like to print off their top 100 at the end of the college football season and then compare it to their final revision.
One player that really caught my eye this year is Dan Williams. I do a lot of SEC games and in my mind Dan Williams is a late second to early third round pick based on production on the field. At the end of the season GBN had him at #90, which is even lower than my ranking of Williams. Now they have him well up into the first round.
So what changed? Other mocks? Buzz? A good off season?
If the GBN would stay true to the tape, i think his website would be more creditable, because at the end of the season, he has a lot of players pegged correctly based on producton on the field.