Feel the same way about Widgets in general - but they suffer in the realm of dynamic UI creation. We simply cannot create lots of them on the fly. Hiding and showing pre generated UI gets us so far, but if we rely on this for too many variations of UI, invariably will end up with 100,000 nodes + in a network for building simple column layout style UI’s.
There are only a few things that the parameter COMP is missing that IMO would make it viable for production ready replacement for UI fitting the common column layout scheme. It’s already lightning fast, already custom param based, supports pages, separators, filtering, etc. However I realize these may not be easy things to add or change.
Ability to hide/disable the + icon on the left for expanding the parameter. This is useful for programming, but maybe not often as much for production / final usage in perform mode.
Restrict/allow ability to see or change binding, expression, parmode, etc. It could be useful to allow expression setting to users in some params, but maybe not binding. or maybe none of them for control only UI.
Being able to set things like font, gfx font size, row height, sep height, label position, knob height, colors of sliders, edge, highlight etc. would be helpful as @alphamoonbase said! Maybe styling info in general could be generalized and abstracted away into an optionally provided STYLE dat similar to CSS? This would allow select dat networks for global style changes.
Depending on how flexible styling was made, one could even consider the alternate use case of a single parameter comp per UI element, for building performance layouts and what not…
Anyways, just wanted to throw in my two cents to this post - realize these asks are potentially huge and time consuming on the back end. I also realize that Touch’s UI may be so fast because of how streamlined it is, lack of options etc. The parameter comp just has so much potential if a few things about it could be more configurable !