Linux is really important, but it's the best-funded FOSS project already. No other project gets a comparable amount of time by comparably skilled developers. A

Linux is really important, but it's the best-funded FOSS project already. No other project gets a comparable amount of time by comparably skilled developers. An open source driver for any particular piece of hardware is only useful for that exact piece of hardware, and maybe a few generations after it. But it's usefulness dies with the usage of the hardware. Contrast with basic research, which stays around forever and can be meaningfully built on for much longer. Research is more about underlying concepts than about actually building concrete implementations (be they software or whatnot) (though that may be part of the research process - but only as the means to an end).