Wow, I love the visual rendering of the schedules; it really shows where the gaps would remain. Now I want 3 trainsets and 10 round-trips!
GO'K, I wasn't considering an additional RT in the total cost equation, only the fact that running faster is more expensive for fuel, equipment, and track.
Once you have the schedule "string board" set up (I just made this one on PPT), it's quite easy to move around the EB and WB movements to see how they can fit. Obviously, it doesn't take into consideration intermediate stops/ route speed variances, track occupancy/slot restrictions on/at MBTA/N Station, etc. but it's a cool little tool to play around with and well illustrates the challenges facing NNEPRA in optimizing their schedule to balance trainset cost/utilization and time-based passenger demand. A third rush hour (or pre-rush) train is just impossible without a third train set... and it hit me when doing this that the service only provides one inbound rush hour train, when they could theoretically offer two.