Very nice indeed !
I am pretty sure that we can do the same in C4D since Bodypaint exists, by painting alpha texture of a multi layers shader.
The textures principle did not evoluate indeed, I think it's because of the problem of the texture itself, which is simply a 2D image, so we have to unwrap it on a 3d model and I don't see how to change it.
Point vertex color become more populated since graphic cards can handle a huge amount of vertices with no problem, more vertices we have, more detail we can add and can replace a texture in some conditions, it's the basis of the shader computation and it's used for always in photogrammetry and laser scanning and radiography.
A revolution would be kind of 3D textures scanned from the environment, the texture will be a cube instead of a square, with depth informations, and be applied like procedural C4D shader (and its wonderfull wood shader), so no unwrap nor UVs, that would be very nice !
I imagine the 3D texture can be a very complex procedural texture, with algorithms stored instead of pixels (hey but the JPG compression isn't an algorithm based on 8 pixels squares... ?), or it can be a huge box of voxels, or both ?
I don't think that a 3D texture could be scanned from the reality (because we can't see through walls), but I am sure that an IA could extrapolate a 2D texture into a 3D one, the 3D texture will be a multiple layers of 2D textures with logical continuity in depth, imagine a 3D brick texture that looks like a box of 3 cubic meters full of bricks.
Some compressions algo like a "3D JPEG" can be invented to redure the texture weight.
I hope it's not too messy, it was just an idea.