There are two main problems with using ground radar sources for mlat:
- Interrogations are on 1030MHz which is too far from 1090MHz for a dongle to see both at the same time. (Also, the 1030MHz modulation is probably too fast for the dongle’s sample rate)
- Ground radar is on the ground, so not many receivers will have line of sight to the transmitter.
Mode-S-only test beacons (commonly “TEST1234”) could be used for synchronization in addition to ADS-B messages if their location was known, but ADS-B aircraft already work pretty well for that and ground beacons still have the line of sight problem; and I don’t know where you’re going to find a reliable, updated, list of beacon positions anyway.
Fixed beacons that transmit an ADS-B position should already be being used.
These things could add extra clock synchronization sources, but clock synchronization itself is not the limiting factor for mlat based on rtlsdr dongles, we already get synchronization down to about 500 nanoseconds. The limiting factors are the sample rate, clock stability, and the dongles tending to randomly drop samples; those are all limitations of the hardware more than anything.