Full disclosure: having become pretty disillusioned and disappointed with the general development of openHAB, and OH3 in particular, when I tested it with looking to upgrade when it came out, and openHAB GitHub issues tracking and addressing, or the lack thereof, I’m still running OH2 here, and with realising that it’s best to rely as little as possible on particular bindings, I very early on turned to MQTT compatible implementation for all the Govees, Inkbirds, Qingpings, Xiaomis, SwitchBots, IR Trotec air conditioning units, 433MHz and 868MHZ rtl_433 weather stations and thermometers, MiLights/MiBoxers, RF Funksteckdosen and so on.
With everything manually textually defined, as I like to do it.
And thoroughly enjoying the stable long uptime, without any dot release update dramas 
This MQTT versatility and independence of proprietary bindings also brought me to the OpenMQTTGatewat/Theengs project, actively contributing as best I can.
With that in mind, and the fact that the majority of home automation users these relying on auto-discovery I keep testing out discoveries for Bluetooth and rtl_433 devices regularly with different home automation systems, also implementing a requested feature for FHEM users, responding to Jeedom users requesting new features,as they do not support discovery, and making sure that HA discovery works as smoothly and nicely as possible.
My last OH4 test resulted in this
With no-one responding I assumed that proper HA MQTT discovery is not a high priority.
Feel free to continue the conversation there, if you think it might be worth it for me to try again with a later 4 dot release. It might get me out of my OH resignation a bit 