I sometimes find it frustrating that people justify a history or traditional method of doing things as a reason for why it is best. There are many historical events that we now find repulsive, Japanese internment, slavery, etc. Yet some PI's choose to run their labs as though it were the 18th century. If they are famous and well established, they may still be able to recruit a cohort of students that is willing to put up with harsh conditions in exchange for working in Dr. X's lab. However, there is a growing subset of the student population that will no longer tolerate these conditions. It is Dr. X's prerogative to choose how he runs his lab, by and large. However, it does become an issue when Dr. X want to eliminate major part of the graduate program to fit his methods. As many graduate programs in engineering and the sciences incorporate research rotations prior thesis mentor selection (as is common in the life sciences) this starts to become an issue, when students get to see first hand what life is like in Dr. X's lab before they officially join the lab. I can't argue with Dr. X's success, which is undisputed, but I wonder in 10 years if there will be anyone left willing to work in the sweat shop.
I take a different approach. I focus on results and progress. I want students who are self motivated and work toward a goal. I don't like to micromanage and everyone is happier when I don't. (I get a little high strung when I am in micro mode, just as my first grad student.) As long as you are making progress and communicating problems, I'm happy to let you run loose. I don't worry about your hours (my students often are using equipment on other parts of campus). I try to lay out deadlines ahead of time (reasonably) and then let you figure out when to get it done. If you choose to go on vacation for 1 week during the 2 week period, you might be working some late nights to finish. But that is your choice, as an adult. I also encourage students to plan their own calendars for thesis writing and determine a reasonable defense date. It generally works. My lab is more laid back, but the students still publish a 3-5 papers from their thesis work. It's not for everyone and there have been students that switched labs or quit graduate school. But all in all I works, and it fosters the type of people I would like to have as future faculty colleagues. In general, it may be less efficient than micromanaging, but in the end getting a PhD is about learning to be an independent researcher, not a work horse for the PI. I used to think I lived in a department of people who agreed with this statement, but after some recent discussions, I'm not so sure any more. Your thoughts on lab cultures and mentoring styles and how to shape those with positive departmental policies?