Let me make a few suggestions for the DB design.
"Tested Resolutions" and "Tested Aspect Ratios" should be different tables, and "screen_change" should be in the latter. We don't extensively test each and every resolution - we just check to see if they work or not. Also, I don't understand what your "group" and "url" fields are there for.
Under the "detailed report" table, these fields should be removed:
calculated_grade
potential_grade
major_omissions
minor_omissions
cce
All of those are derived values, determined by looking at the data that is (or is not) in the DB.
Also if current DRs was to be put into the database they would probably have to be done by hand, via forms created, this would be an extremely lengthy task
I've been doing more or less exactly that, ever since I wrote the first DR. Granted, it's not perfect - I've set it up more for easy data entry than for performance (the first version was an Excel spreadsheet!). But the data is all there, and it can be migrated programmatically.
The real time consumer, as I see it, would be creating the forms, for data entry, retrieval, and especially administration. For this reason, I've always seen a DB as something that would be nice to have, but probably won't happen. Such a thing would indeed save lots of manpower in the long run, but there's always the matter of initial time investment.