So let's work on a description of how a Blind Cache solution could look like. We're not yet 100% sure if the value of BC is motivating the cost and risk for user privacy and security (won't be as secure as end-to-end HTTPS it would seem but better than h2c and SRI perhaps?).
With a description that can be discussed in the community, a prototype to get down to the gory detail and a nice test bed, we can explore this and find an optimal solution. This we then evaluate and make go/no-go decision.