It seems like there is no modern meta-search engine that gets all the available results (by paginating) from its sources, and then crawls them and re-ranks them. As far as I can tell from the paper, this is what the original meta-crawler did.

My questions: Do you know of any modern ones that do this? Do you remember any old ones that did this besides meta-crawler?

It might be slower, but I think that something like this would be fantastic, especially if it were open source so that I could come up with and experiment with my own ranking methods.