Zen Faulkes: "Does a Ph.D. train you to head a lab?"
The big one, though, is bookkeeping and budgeting. I didn’t have to worry about tracking money in any significant way as a grad student or post-doc. Spending money at an institution is not like spending your own money. You have layers of people and paperwork that stand between you and purchases. You have obscure “enterprise finance” systems that seem designed to drive a person to substance abuse. I’ve learned that I despise trying to keep track of grant money.
He discusses many other ways that a Ph.D. course fails to prepare students for research independence. I have an orthogonal thought.
Running independent research requires entrepreneurship. A Ph.D. formally is training in scholarship. A great scholar may be a poor entrepreneur, and few Ph.D. programs require training that would instill values of entrepreneurship. Essential skills include professional networking, balancing risk by diversification, repeatedly and widely asking for funding, accurately judging the motivations of people who share information, and publicizing and promoting one's own ideas. Some students get excellent informal training in these skills, but many miss them entirely.
Being both an honest scholar and a successful entrepreneur requires a very special package of talents, and taking on research independence is a difficult transition for even the best students.