Assuming you're right that this is a trend, I don't think it should be too hard to heuristically pick out likely examples. Assuming most of them are good quality, there should not be so many that it's impossible to pick through the results and do the final formatting manually. Waiting for a standard format to come along could take a while (seen the semantic web yet?).

The jekyll-scholar format already exists and is more "semantic" than Markdown. After people read these database papers, perhaps one will understand the value of indexing structured data and convert/fork the list repo to structure the source format. This is easy at the birth of a collaboratively edited project and impossible later.

> Assuming you're right that this is a trend

Here is one list of lists, https://github.com/sindresorhus/awesome

My experience has been that expecting people to do more work, even a little bit more, doesn't often pan out.

At least MS/Google will thank humans for enriching their proprietary knowledge graphs / semantic webs. It may take a few more years for people to realize why they need "personal algos" as much as "personal computing devices". Humans can then feed delectable metadata to their private algos, in addition to feeding large public spiders.