The experiment seems ripe for gaming though - just have the control group gradually lose productivity (if needed) to be just below the test group, so it seems like 30hrs/week is definitely better than 40hrs/week. Assuming they're measuring productivity with some easy to game number, which I hope they aren't.
You need a pretty long term experiment, so the 30 hour group gets the chance to settle in their new rhythm, and you need a control group that doesn't know they're the control group, presumably from a different city that happens to have very similar productivity.
The interesting part of the analysis is that obviously a 8 hour workplace can handle a fraction of the employees calling in sick pretty much transparently. Now what happens when everyone works 6 and there theres no slack WRT sick days?
From personal experience the time and monetary savings of four tens is far more appealing to me than the idea of spending perhaps nearly a quarter of my time away from home, going to / from work. As a starving student I worked 4 hour part time shifts working while going to school... at $1/gallon gas it barely made sense, now gas being permanently around $4/gallon I don't think short shifts make sense unless your commute is extremely short or you make a substantial amount of money per hour.
Well, actually they get the same thing they always got, and if the experiment goes well, they may get more free time later, but some of their lucky co-workers get the extra free time right now.