it might even be true that visual/positional reasoning is necessary in any formal language

It's definitely not true that visual reasoning is necessary for programming/formal languages, since blind people can program just fine. This is one way in which it seems to me more visual approaches to programming are a step backward even if/when they work - they make it fundamentally less accessible to blind folks. On the other hand, would we deny engineers the right to use diagrams in order to make engineering more accessible to the blind? Probably not. And it's not as if we don't often use diagrams or visual notation in explaining our code already. So, yeah, I'm conflicted. I'm not sure what exactly you mean by "positional" reasoning. If you just mean stuff coming before/after other stuff, sure, that seems necessary, or at least I can't see how to do without it 😛.

Hmm, it's actually a good question whether vision is necessary for visual programming. If so that seems a knock against it on accessibility/moral grounds. But I think the answer is fortunately 'no'. I think most blind people may have a pretty well-developed spatial sense, and that's really what visual programming is trying to tap into.

Spatial intuition might be what visual programming systems are trying to use (although I don't think that's the whole story), but the way they accomplish this is... well, visual. Maybe in some imagined future they could work in a tactile way, but I think that's so hypothetical it's not useful.

Yeah, it's all hypothetical given how nascent the field is. I'm not invested in visual programming, and like I said accessibility is a key question any proposals have to answer.

this is a really good point! I think spatial reasoning is a better way of putting it —and dynamicland would be a good example in some cases of a more accessible form of spatial reasoning (how cool would it be for them to design a fully physical syntax for realtalk?!)