A 75 page paper usually takes longer than 48 hours for normal students to write. If the version history for the paper is submitted, but all the commits are within a very short time frame, then either the student stayed up all night to write it, or they paid someone else to stay up all night to write it.
And if you ask them about what the wrote on on something they took all night to write, they should know it fairly well. But if they paid someone else to do it, they wouldn't know anything that's on the paper.
So then this big company hires a programmer to write a tool to randomize entries over a week or a month or whatever, and our shadow scholar has 1 additional step when he finishes up.
Basically, this idea is DRM in another guise, and terrible for all the same reasons. If you don't trust the student that much, then simply abandon take-home work in general.
Randomized entries would be obviously fake. To be a convincing fake, they have to make sure it doesn't conflict with the student's work schedule, class schedule, and so on. That means deciding when the student was supposed to have been writing the paper, which requires quite a bit more coordination and drives up the cost.
