All they are saying there is that if you want to prune the signature data, you need to still keep a hash of the signature data in the chain of hashes (of Merkle trees) for the blocks. In other words, you need to still be able to prove which signature signed which transaction, even if you've actually discarded the signature data.

I believe BBR already does the correct thing. And afaik, Monero does not discard signature data, but I could be wrong about that. If they do, I assume they would do the right thing as well.

Any way, I as I read into the proposals more, I realized they are adopting some of the ideas I've had privately, but bolting these onto Bitcoin's legacy limits their flexibility in terms of optimum choices and especially speed to market. Bitcoin is looking more and more like design Rigor Mortis (they even need to abandon ECDSA to totally rectify malleability) and needs to be scrapped and start over again.

Wuille wrote this technical summary in way that can only really be understood by other core devs or experts who have their head deep in these issues. He sounds rushed.