Stuff I've tried is mostly about talking, trying different systematic models (time tracking, semi-micro management, constant reporting or leaving them more space, trying to get excited with new technical bits, giving them bonus, cutting their salary etc., pep talk, pissed-off talk)

Many things, not done in exact science although it was obvious what slightly works and what definitely doesn't.

> Also it is hard to work remotely and many people just can't do that.

That's exactly my point, maybe these 5 guys can't handle remoting but I might find new 5 guys who can (just like I already got 3)