You raise a good point. Little overhead, the endpoints are well-formatted, you can get a digest of articles in one blow without API keys, and you just need to parse the resulting XML.

That seems reasonable, although there is no telling what a highly voted post might constitute under new management (is that too paranoid?). I’d personally take a scrap and build approach here, or at least manually approve the incoming results (hybrid approach) if they’re being delayed anyway due to waiting on vote generation.

In practice it’s not so easy without some manual curation. News sites post a lot of filler stuff and you don’t want to start spamming yourself with every article posted to <insert magazine here>. Even on higher-traffic subs you don’t generally see more than one or two posts from the same site on a given day. It’s generally more effective with something repeatable and reliable like a weekly column where the expected “quality” is invariate. Certainly you can front-load the manual curation by building a set of filters into your scraper, but whether you filter the results at the front or the end of the pipe, you still need some kind of heuristic for what constitutes “good” content, and that’s frequently a moving target.

It’s trivially easy, but you’d be pulling in a lot of noise along with the signal, creating more moderation headaches for yourself (think of all the low-effort and spam stuff you usually have to filter out). You’d be better off scraping the content you want from primary sources directly rather than mirroring every post that goes to your old forum.