Not sure about autonomous, but just something you own, that you have full schematics and design specs for, and can repair yourself — seems like a huge market given the state of things right now.
It's true that parts availability has been somewhat of a problem as of late, especially when computer chips are involved, but I am not sure it being open source really improves on that. If the OEM cannot keep up with production, your DIY/local supplier efforts are likely to suffer from the same capacity problems. It's not like you're going to setup your own silicon fab on a whim when your tractor breaks down in the middle of harvest.
There is definitely demand for tractors from new producers. The inability to repair a tractor during harvest is simply unacceptable. Their technicians can take weeks to get around to you, especially if you are a small shop.
Asking out of plain curiosity.
Marcin et al., had also designed and built an open source earth brick maker, and trialed it in building buildings out of it. The advantage is being able to source building materials on site.
I don’t know how far the organization has gone in designing and prototyping the other devices they have on their roadmap.
I think people who aren't involved with farming can't really comprehend the amount of technology involved and the savings that technology enables (and ultimately makes it worth it even if the equipment is more difficult/expensive to repair).
As an aside, we always hear complaints about John Deere and never other manufacturers. I wonder if it's because anti-corporate populism works better when it's targeted at a well known american brand or if John Deere is truely that bad. We don't have any modern John Deeres in my area but a few Case IH and Claas, and I've never heard any complaints.
Btw my point isn't to justify the anti-consumer behavior of some manufacturers, but to provide context as to why "an open source tractor" is a completely naive idea.