My guess would be to take a look at some heuristics on the relative project size. Perhaps number of use cases and the complexity of those use cases? Or test scenarios.... something that is 'known fact' about the project which you can then place some 'value' on the relative size/weight. Other areas may be the number of resources, project budget, etc....
Other items to consider is the Project Management style used on a project. I'm just throwing stuff out here - but - if the average project manager operates like this...
1) Holds 30 minute standing meetings each day to coordinate team activitites
2) Holds 60 minute status meeting each week to assess progress
3) Holds 60 minute steering committee meeing each week to report progress, issues..etc.
4) Spends 60 minutes a day updating issues, risks, handling calls...influencing others.
5) Spends 120 minutes a week prepping for #1, #2, #3
Well - right there- you have about 8 hours of PM work on a single project. Assuming that the project is clicking along well... ;)