really nothing to add to @apella12 summarization.
I started with OH back in 2017 because of ZW binding and from my point of view ease of use compared to HA back then.
After deciding for OH and defining my use-cases, I did hit a few issues with OH + ZW binding (Z-Wave Binding: Fibaro FGR-222 Rollershutter UP command not fully raising Blind and Fibaro FGR222 Roller Shutter 2 in Venetian Blind mode - use of manual switch no updates sent to OH2.4), but was too deep into OH and finally accepted those 2 issues as “known limitations” (mainly to due to lack of time to migrated all to HA).
After learning about Zwave-JS-UI and finding some time around Christmas to investigate and test the migration to Zwave-JS-UI I was convienced that Zwave-JS-UI is the way to go.
The number of devices supported by Zwave-JS-UI is another compelling advantage over Zwave binding, refer to Zwave-JS supported devices, and so is the larger user base.
In my view the switch to Zwave-JS-UI makes sense for all network sizes, even if MQTT needs to be introduced at same time.
The migration from binding to Zwave-JS-UI was as simple and as straight-forward as described in the previous post.