If you asked me to bet at even odds on who would be able to do two, or five, or ten launches to orbit first, my bet would be on Starship.
If you design a mission specifically for SLS then yeah, SLS difference matters, but if you simply want to make a good architecture, you much rather go with the smaller cheaper rocket.
In terms of Orion, can Dragon do everything it can do. No not by itself Orion plus the Orion service module can do some thing Dragon can't. But Dragon can do most things and with a few simple add-ons like extend tank in the trunk.
So in reality if NASA had the freedom to plan, a few years ago they could have switched a Falcon Heavy/Dragon architecture and achieved the same thing. SpaceX would likely even have been willing to do things like cross feed and potentially even a Raptor based upper stage. All of these would have been ready far earlier then when Orion will actually launch with humans.
So just saying, SpaceX doesn't support something is a really narrow view. It doesn't support it but SpaceX certainty would have the capability to do the Artemis 1/2 missions if NASA had asked them too.
And all of that while completely ignoring Starship, just with the capabilities they had in 2019. Between then and now SpaceX could have already done this mission quite easily. And likely they could have done it 5 more times before Artemis 2 launches.
If the private launch services didn't exist how they are, SLS would be heralded as the latest generation of rocket.
The timing is terrible for all sorts of reasons.