Generally speaking, though, accuracy of the sensor is rarely the limiting factor in gravity surveys.
For land-surveys, it's precisely knowing your elevation, correcting for the "unwanted" mass distribution around you, etc. For mobile surveys, it's correcting for the acceleration of whatever vessel the instrument is on. Any ideas you might have for improving the state of the art for the mobile case would probably be _very_ marketable.
On a separate note, though, the inverse problem, while fundementally very non-unique, is still solvable for many practical problems (e.g. we know the range of density of the materials involved and we can make a reasonable starting guess for the distribution of mass). Regardless, you're usually interested in distinguishing between a few scenarios that can be easily forward-modeled.
A lot of the imaging market appears to come from security/defense applications, either in portal-monitoring or for IED detection. There are defense contractors working on both. The former is easily spoofed (put your uranium pit in a styrofoam sphere in a truck full of grain, done), and the latter is hard to do at speed in a rugged environment.
I've spent a lot of time trying to figure out how to do gravitational imaging on the sub-meter scale, and while it does work, it's hard to get sufficient image resolution to be useful for anything other than a party trick.
If our other science weren't more interesting, I'd be doing it for fun alone. Burning 3-6 months on the project to assess feasibility was as far as I wanted to go without a clear exit strategy.
I wonder, however, if the problems you are describing are things that would work themselves out over time as the equipment improves? The precision of these clocks sound like they're many orders of magnitude better than the old gravity sensors. I think. Perhaps all we'er waiting on is some kind of crazy technological magic over the next 50 years that would involve miniaturization, improved accuracy, and an array of a million or so. (All of which I just made up)
To do it with a pair of clocks, there's a long way to go, as they measure differences in depth in a gravitational potential. If you change your distance from Earth's center by a meter, your gravitational potential will change by 9.8 m^2/s^2. If somebody heavy (100 kg) and spherical (I'm a physicist) stands a meter away from you, your potential will change by 7 x 10^-9 m^2/s^2. In short, the clocks need to get ten million (they can see a 1 cm height change) times better to sense a nearby person. Clever trickery with an ensemble of clocks will make that easier, but not by 10^7.
Not impossible, just hard. The fact that the time/frequency teams have encountered gravity, in particular through the gravitational potential, has just made their lives quite a bit harder. They now need to know a lot about the relative locations of other clocks and the mass distribution within the earth to make substantial strides forward. It's an incredible feat to have gotten to where they are, and a major challenge for the future.
If there's any application for pure gravitational sensing that can resolve the position of a higher-density mass to a few centimeters over distances of perhaps a meter or that can make simple statements about a meter-scale mass distribution, please drop me an email. We've tabled the project because we don't know of a single use for it, academic or commercial. Device cost would be in the low hundreds of thousands of dollars and require careful operation. We've thought hard about this, but haven't ever found an application for which some other sensor wouldn't be far more appropriate. X-rays, neutrons, resistivity, clever weighing, optical techniques, microwaves, touch probes, three-year-olds, lemurs, optical imaging, you name it, it's probably cheaper, better, and faster.