Why I expect SMS involvement in WRC game is that EGO engine does not do all that Codies have wanted to achieve in rally simulation. Since DR1.0 they have been talking about shifting conditions for drivers, and how the games are still too far away from reality in that. While DR2.0 has some attempt to it, the limits of the current engine prevent them from going further, unless EGO is developed much more.
We know that Livetrack 3.0 already calculates surface temperatures so that areas in the shadows are colder than areas in sunlight, water flows, dries up etc. By now I expect that Livetrack can also calculate water freezing and melting in those areas too (was not in PCARS2 yet). It would finally be possible to do Monte Carlo properly without the need for baked-in shadow maps and manually setting up the conditions inch by inch.
For the gravel tracks Livetrack also simulates sweeping dirt on the surface, again dynamically, so you can simulate any starting order in rally.
At least currently EGO engine does none of these. Properly modeled rally stages are longer and much more evolving than circuit stages, and manually doing this kind of stuff gets cumbersome really fast. For instance it took months for Codies to release more day and weather settings for some of their existing stages.
And currently in DR2.0 there is not much point in variations in the conditions of Monte after you learn the spots where there is ice or snow; they never change.
I also expect that all stages in new WRC game will be real ones, and SMS already has lots of experience with photogrammetry from PCARS2, and PCARS3 too I suppose.
Apart from just developing tyre model from circuit racing perspective, SMS has put incredibly amount of effort for creating this versatile engine from the scratch. For the players snow stages in PCARS2 have not made much sense, but for SMS it has been there because they needed to develop it anyway for their grander vision. PCARS2 was just a good opportunity to showcase it to everybody in the business on the same go.