Application pipeline - one or more fact tables?

Hello all,

I am working on modelling our  grant application process. The process involves several distinct phases, namely application receival - qualification/disqualification based on formal eligibility (application form filled completely eg) - evaluation - grant approval or disapproval - contract

For me, it seems that accumulating snapshot - one row for each application and 1/0 columns and corresponding dates to indicate the movement of the application in pipeline - would model our process nicely. My question is this: would it be better to model the process as transaction fact table and then derive the accumulating snapshot from that table? (I have understood that there are some disadvantages with accumulating snapshots when it comes to counting sums for given time periods?)

Another question: If we decide to go with transactional fact table, would there be some advantages associated with segregating this process into separate fact tables (one for application receivals, one for evaluations, one for approvals/disapprovals etc.) or would it be better to just load all the transactions in one big fact tables?

Thanks for your help!

Antti