You can use passive radar for surveillance radars, but is it any good for tracking and guidance radars? If your radar stations can see F-35s, but your interceptors and SAM batteries can't fire at it, the stealth is still useful.

Stealth is relative; as long as your detection distance is greater than the adversary's, you're good. Also, passive is great, but I'd expect in a war situation all large transmitters would get disabled first.

All cellphone towers?

You don't have to worry about most small cell towers, since they are usually directional (pointed towards the ground to maximize signal coverage) and relatively low power. In the countryside with tall towers that cover tens of kilometers... well... it's kind of hard to miss those with a missile.

You can always mount the transmitters in trucks and keep moving. If you can pinpoint the source and fuse the input of, say, dozens of small sensors deployed all over the place, you can pinpoint the plane well enough.

Not well enough to do reliably do missile tracking, just enough to get an idea that yes, there is an object out there in the atmosphere. "Dumb" methods like TDOA / FDOA have quite bad limitations on resolution.

Multistatic passive radar is quite hard to do at high performance. Radio astronomy observatories have been doing it for decades; it's expensive, requires high grade equipment to synchronize time and frequency, and reliable, real-time high-bandwidth data links.

That's getting awfully ahead of things. Showing that it's sometimes possible to detect aircraft like this is a long, long way from demonstrating that it's even theoretically possible to replace all existing air defense and targeting/interception radars with it under all circumstance, much less actually trying to do it.