Remote mass detection and location Could a collection of gravitational sensors (eg. MEMS accelerometers), placed in a 3d grid (eg. 10 x 10 x 10 sensors forming a 1 x 1 x 1 m cube) detect the the presence and position of a remote mass concentration, eg. an airplane in the sky?
What accuracy would be needed to do so?
 A: Your accelerometers would be measuring acoustic waves from the airplanes and the phenomena.  There is already a global array of infrasound detectors and seismometers to monitor for nuclear tests, and their sensitivity is exceptional enough to detect things like plane crashes and probably even cruising planes and trucks driving
As far as monitoring the gravity disturbances of these things, we've only just developed the technology to detect large black hole collisions, events which produce many times the combined power output of all light from all the stars and galaxies we can see. So it's safe to say detecting the gravity of airplanes is out of reach for now.
A: Gravitational sensors in orbiting satellites have been used to map density variations in the rocks under the ground. (See here for a recent example.)
At shorter range, gravimetric gradiometers are being researched for detecting underground tunnels, objects behind walls, and stealthed aircraft and submarines. (See here for a list of recent projects.)
You would probably not have much success with a 1x1x1 metre grid; you need to take measurements across a wide range of angles to localise sources. So you either have to move your sensor around (the usual technique), or have a widely-dispersed array of sensors.
