There are actually things to test in the integration between the rocket and satellite. It's not like the satellite is a dump piece of steel sitting on top of the rocket. The payload has its own computer, power systems, and propulsion. The interactions between that and the launching rocket / ground systems matter. Also mating the two is non-trivial, so you can push the static fire back closer to the launch if you are doing it with the payload on board.

A static fire without the payload on the rocket was already done in Texas several weeks ago. This is the final dress rehearsal before launch. I always thought of these tests as low risk, but I guess nothing in rocketry is low risk.

Would they do so if human cargo?

Actually they specifically plan to load astronauts before fueling begins. If they're in the capsule the launch escape system (LES)[1] can propel them away from any on-pad disaster; not so if they're still walking across the gantry.

That said, NASA has raised concerns about loading crew pre-fueling as part of the NASA-SpaceX Commercial Crew development contract (CCDev).

[1] Here's a video of the LES test for Dragon 2: https://www.youtube.com/watch?v=1_FXVjf46T8

Note that a human doesn't "plug in" to the rest of the launch vehicle the way a satellite does.

I imagine they _would_ do these tests with the crew compartment attached and dummy human loads strapped in to the crew seats.

That is how Virgil I. "Gus" Grissom, Edward H. White II, and Roger B. Chaffee died.

https://en.wikipedia.org/wiki/Apollo_1

They have a launch escape system. Although even when it works the human cargo might need cigarettes and "shots of vodka to help them relax"

https://en.wikipedia.org/wiki/Soyuz_7K-ST_No._16L