However, they tend to be finicky to work with--difficult to calibrate and to eliminate sources of error. After all, we know from Einstein that there is no measurable difference between gravitational acceleration and inertial acceleration (like, say, bumping a sensor).
Still, gravity seems like an area of immense promise for sensors of the future. We've spent immense effort improving and miniaturizing sensors for things like magnetic fields, electrons, and photons. But we've recently discovered--using gravity--that these things might represent only 5% of the total mass of the universe. We haven't even detected a gravity wave yet. We're just barely getting started at using gravity to observe and understand the universe.