Is there a better benchmark to use?

Honest question.

How can a lay person track the real word progress of quantum computers?

Most approaches have missing "capabilities" that can be tracked. Adam Zalcman lays them out for superconducting qubits here. https://westerbaan.name/~bas/rwpqc2026/adam.pdf

For the neutral atoms approach in particular there doesn't seem to be a clear capability missing anymore to building a full scale CRQC: each of the separate components has been demonstrated. Of course when they try to put everything together they'll undoubtedly hit unexpected issues with integration. Wish I could be a fly on the wall at those labs.