This is going in a direction that manufacturers have so far mostly avoided probably out of tradition.
These days lens design is already highly automated with simulation software to do the optimization. High-end lenses are just lenses where you've relaxed some of the conditions to get better quality (bigger, heavier, expensive types of glass, lower manufacturing tolerances). And yet some of the parameters lenses are still optimized for (color transmission, field curvature, etc) are increasingly something you could fix in post-production if you had an accurate model of the lens.
This makes particular sense for non-interchangeable lens cameras where the sensor+lens combo is known and manufacturing tolerances are already low, so your lens modeling can be quite good. It should be particularly useful in smartphones where all the other restrictions are so tight (small size and weight require it).
This is already in use in the micro-4/3rds cameras, which feature interchangeable lenses, but still tight integration between the lens and the camera. There are even firmware updates for the lens.
Neat! I had seen it done on compacts but didn't realize anyone had done in on interchangeable cameras. There are plenty of lens correction packages for DSLR lenses but those are "since the lens has compromises we fix it programatically" not "the lens was designed with the correction in mind".
Major manufacturers are indeed adopting these techniques for correcting certain flaws. Recent Canon DSLRs will correct for vignetting and chromatic aberration using lens-specific profiles.
Those were the cases I knew where a manufacturer goes and says "here's an electronic recipe to fix some of the issues we couldn't on the lens", you're still getting the uncorrected photo from the camera and then choosing to fix something in post, but the lens is supposed to work to a high standard without any fixing.
The example ErsatzVerkehr mentioned in m43 and is done in some compacts is more "we have a lens+sensor+software combo that spits out a correct photo and you never see uncorrected results". The difference is that you can design lenses that without correction produce unacceptable results.
These days lens design is already highly automated with simulation software to do the optimization. High-end lenses are just lenses where you've relaxed some of the conditions to get better quality (bigger, heavier, expensive types of glass, lower manufacturing tolerances). And yet some of the parameters lenses are still optimized for (color transmission, field curvature, etc) are increasingly something you could fix in post-production if you had an accurate model of the lens.
This makes particular sense for non-interchangeable lens cameras where the sensor+lens combo is known and manufacturing tolerances are already low, so your lens modeling can be quite good. It should be particularly useful in smartphones where all the other restrictions are so tight (small size and weight require it).