I think this reasoning doesnt depict the real model.
If you are supposed to do 15 hours of real work, the ambiance of leaving when you are done with it is in the best case, going to let you do those 15 hours, but worst case you wont even fullfill that.
An example of that is finishing up something, and then a bug pops out after someone checks it out. If you are not in the office, your work is incomplete.
For the value vs work proposition, its entirely inapplicable for everyday operations. Who is measuring the value? and how do you get to influence those decisions?
If you get to work a project brought down from the product manager, you implement it completely, and the project turns up to be a total bust, or just useless, or not used, you provided 0 value. Does that mean you have to work extra to make up for that?
Or you work a couple of hours rogue-like and discover a dev-ops inefficiency and you can cut down server costs 10%, saving the company 100ks of money. Does that mean you don't have to show up for work anymore?
So value is not the word at play here, its maybe the efficacy of going through your work load. The more time you spend commited the more you will do, and that is pretty strict. The guy doing 15 is not going to get more than his own version doing 40, no matter what.