Trying to measure coder productivity with any of these metrics is almost a waste of time. If metrics like these are the criteria, then any humans will tend to game the system.
You want me to write more lines of code? I will write more lines of code. You may not like what those lines do, though.
You want me to fix more bugs? I will fix more bugs, but they'll be the easy bugs. No one will touch the hard bugs, if they each take six weeks to fix but only give credit for one bug fixed.
I'm not saying anyone would be uncooperative or anything -- far from it. Just that people want to do a good job, and if the definition of a good job is to meet certain metrics, then they will do their best to meet those metrics.
Joel Spolsky included an article about this topic in one of his books. Here it is:
Larry's rules of software engineering #2: Measuring testers by test metrics doesn't. by Larry Osterman, April 2004.
Would I quit a job at such a company? I like to finish what I start, so I would not quit immediately. However, I would ignore their arbitrary metrics, and work with a level of quality that I feel is appropriate. I like to do a good job, and generally my managers have been happy with my work, without using a lot of metrics.