I think I've got 2 points to make about the replies, the latter also being about where I was trying to get the conversation going. All are linked, so let's have a go.
1. Assumptions about Infrastructure being the main effort mirror the road safety establishment as it was from the 1920s ref car safety. Their take was the 3Es, Education, Enforcement, and Engineering. First 2 were about the driver, and the third about the road. We do bang on about those, and they have their place for sure. But we ignore an aspect [bike design] we might be able to improve on, no?
2. Not sure the bike can protect a person from a 50mph side-swipe, but there are various common 'failure modes' of when and how people come off bikes or hit/are hit by something. I've had my fill on DF offs and they nearly all have a common outcome of trying to avoid head on concrete from height. Bents have different angles and shapes to deal with, so the scientific rigour could well be applied there with the right historical data feeding the tests. We might end up somewhere totally different to either, that's the fun of it.