This is not a new idea at all. Most of that makes sense and indeed that is the argument for going to "scheduled" railroads. The problem comes with the disconnect between theory and reality. You slip that one phrase in
and have them stick to them
If it was only that easy! When a train pulls a drawbar, the power dies, the train ahead hits a auto at a crossing, the track gang takes a track out of service, a conveyor breaks at a power plant unloading a unit train, a heavy snow hits....well, it could fill the entire page but you get the idea. Any of those things can make the schedule go out the window. And if it's a train which gets handed off to another crew, now THAT crew has to start later, or else spend half their shift sitting on their butts waiting for the train (and probably outlawing before they get to THEIR terminal).
Another factor is that carloads are not constant. As shipments ebb and flow, a branch line may generate 15 cars one day and 55 the next - or even more.
A scheduled railroad can be desired and planned for - CN has done a lot on that concept - but there still will always be a degree of uncertainty. It's not as easy as it seems, and some times it makes good business sense to deviate from the schedule. Your train from A to B is SCHEDULED to leave at 8AM each day. Today's train has 15 cars, but at 11:30 another train will arrive with 85 cars to go from A to B. Do you run one train at 8AM with only 15 cars, then call an extra for noon for the other 85? Ooops - "call an extra" - guess THAT's not scheduled, is it? Or do you set back the 8AM crew to noon and take all 100 at once? Or do you let those 85 cars sit till TOMORROW's 8AM train, and hope that another 85 don't show up in the meantime?