[ I know this is probably on the to-do list, but anyway I am noting it here so that it does not get forgotten ]
For those sensor equipped with GPS (E.g. Radarcape) it would be great if the GPS position computed by the GPS receiver would be sent (E.g., periodically) to the database. An important companion data that should be reported along with the GPS position p is the number of satellites Nsat that were used for the fix (this is found in the NMEA string). In this way, one can easily check whether the reported position p can be trusted (Nsat>=4) or not (Nsat < 4), and whether the sensor was relocated or not. Moreover, for static sensors, one can post-process the sequence of GPS positions collected during a long observation interval, e.g. filtering out the most accurate position reports (with higher Nsat) and in this way improve the position estimate of the sensor.
The point I want to make is that both p and Nsat should be reported (ideally), not only p.
For static sensors, one could configure the GPS reporting every X minutes (e.g. every 5-60 min), with the understanding that the the position with highest value of Nsat should be reported (higher Nsat --> higher accuracy).
For moving sensors (perhaps we'll have some in the near future) of course it makes sense to report the GPS position more frequently.