Reading the article it also sounds like lots of poorly designed software. The author complained about pairing being a chore but it seems like with some good software you'd just put up any beacon, scan it, then pick what it corresponds to. If that beacon disappears or breaks just pick any beacon out of sack of beacons, place, scan, pick what it corresponds to. That seems like it should be a trivial 5 second operation. Am I missing something? I shouldn't care what the id is. I only care that it's unique to every other beacon (which the software can tell me the moment I scan it). I don't care about major or minor numbers. Just put any beacon anywhere, scan, pick what it corresponds to. Done.
Note that for school kids, a significant user of museum services, my kid's school issued ipads can scan any QR code and they use that in class, in fact its the only time I've seen someone "in the wild" actually using QR codes, but getting an app installed requires an impressive level of government bureaucracy and leadtime (basically once a year at curriculum review time you could get a new app added to next year's standard system if there's storage space and the review board approves, etc).
They rolled out unlocked, which led to nothing but playing games, then went thru a phase of blacklisting but there's too much whack a mole to list everything to block, now they're pretty well locked down.
That is key, they have to have your app installed (and allow the correct OS permissions). After that you can (on iOS at least) pop up alerts, passively track how people are moving through your space, see what the most popular spaces are, change your app context to show them details about where they're standing, give them directions, etc.