I skimmed the paper. It looks like they eliminate the step-like verification times. That's all. BP verification times are the same when the number of outputs is less than the next highest power of 2. That's why they say "we have faster verification when number of outputs is 9." That's the largest gap when you move from 8 to 16. I don't think this

is very useful because the vast majority of txs are 2out.

and I don't see many 9out transactions. Usually tx with many outputs already max out at 16

<a​ck-j:matrix.org> Agreed. It seems like the results are cherry picked. Not including BP+ or ++ in comparisons and showcasing benefits for minority transactions. However, it may still be productive to invite the authors to discuss their results. They got back to me and said they were busy until friday. Maybe we could have them join the next MRL meeting