Yeah. And the transmitter piece has been broken since Maria hit, so we've already gone quite some time without that capability.

I think fast is also limited to a much lower frequency than Arecibo due to trading off pointability vs surface consistency-- which it needs in part because its so much further from the equator.

That is not good. Was Arecibo the best instrument for early detection of asteroids? What are we using now?

Now, we use optical scanning telescopes with a wide field of view, and image differencing.

Optical instruments have a wide field of view and don't have the inherent SNR problems that radar does. (I.e., using the Sun to illuminate objects is more efficient than doing it yourself.)

As a side-effect, there are all kinds of other transient sources that these scanning telescopes pick up, like supernovae, relativistic jets from AGNs ("blazars"), exotic variable stars, etc. (http://crts.caltech.edu) It's a zoo out there!

Radar does play a role in orbit characterization, and recovery of object geometry, once the NEO gets close enough. But you can get orbits through optical techniques too.

The NASA center for near-Earth objects (NEOs) is: https://cneos.jpl.nasa.gov/about/search_program.html

NEO close approaches to Earth, tabulated: https://cneos.jpl.nasa.gov/ca/. There was a really close one today, actually (a 5-meter object at 0.13 lunar distances).

For more about the role of planetary radar in NEOs, search for "radar" in this FAQ: https://www.nasa.gov/planetarydefense/faq