(context: I used to be involved in the design of military radar systems for the Dutch navy)

The radar absorbing compounds of stealth aircraft are highly optimized for specific wavelengths (usually X-band) and fall off heavily outside that frequency band. Similarly, the radar cross section of stealthy aircraft is highly optimized for specific purposes (usually evading GBAD in the forward direction) and rapidly falls off in other scenarios. Most "stealth" aircraft are actually fairly visible from other directions.

That said, multistatic radar with transmitters-of-opportunity like cell towers and civil radio stations has always been in strong competition with fusion power as "the tech that is forever 10 years in the future". The transmitters are often not very powerful compared to dedicated radar systems and worse, they transmit energy in the horizontal plane rather than upwards where the planes are. The frequencies involved are much lower, which inherently leads to less radial accuracy unless you use VERY large antennas. Unlike a dedicated radar system the signals they send out are typically not shaped optimally for radar purposes, so signal processing like pulse compression becomes much harder. Because the signals are inherently not as predictable as normal radar signals you need MUCH more computing power. Finally, atmospheric conditions become fiendishly tricky for long range, because signal delays between each transmitter-target-receiver triple will be different. This means resolution goes way down if there's too many clouds or ionospheric interference, often to the point of uselessness.

Many of those problems are mostly terrible when trying to detect aircraft at long range though, and largely go away for short range surface use like in port. I'm still not entirely sure why for a port, which is stationary and requires tons of infrastructure investment anyway, this system would be preferable to a normal civilian type radar system. You can get a conventional one for at most a few tens of thousands, while this system apparently requires a trailer full of RF signal processing equipment. That is likely to cost at least in the order of magnitude more, while probably being less accurate.