What you don't seem to understand is there are two variables, site activation, and enrollment. site activation is of course not linear. Enrollment in your simulation, by itself, is linear. you say the combination of the two is linear. I say it is possible that enrollment portion, by itself, can be non-linear. HWINW - WE DONT KNOW WHAT THE PLAN IS!
Quite simply, I don't know what has happened in other trials, but patterns probably exist, and PPHM likely used such historical data to determine the plan. However, I could easily see where the plan could "expect" faster enrollment as time goes on as doctors tell patients about previous patients experiences. So many variables would go into the plan. To simply assume there is no other rationale for anything other than linear enrollment, is a leap I am not about to make.
FFTT
JBAIN