Regarding working in solitude, the study suggesting greater productivity fails to address much bigger factors than mirror neurons. If I'm in a cube, am I even likely to see someone else on their computer. Probably not, so the laboratory effect disappears. And I am more likely to worry about someone (i.e., my boss) seeing what I'm doing, so I'll be less likely to play solitaire. I'm guessing this is a bigger effect. On the other side, I may be more likely to socialize with coworkers, etc.