Yes. And yes. Treaties for arms reduction deal with not only warheads but also delivery vehicles. Protocols include mutual inspection of disabled vehicles, etc.
I have a hard time imagining that anybody with the resources to secure a nuclear warhead in the first place would later be prevented from using it just because it wasn't plug-n-play compatible with their existing trebuchets or whatever.
It's not about "later" it's about "now". You don't inspect standing stockpiles and delivery mechanisms to say "See? We will never have the capability to nuke you", you do so to show that you're not currently building and are not currently sitting on a pile of nukes.
As mentioned probably two posts up, any ramp up would be obvious to adversaries, have some lead time, and would take away from other production efforts.
Are treaties still worth something? In a world where authoritarians defect the "global community" and start wars of conquests that violate MAD , are those papers and laws still worth something?
The "global community" was never a thing. In the fist decade of the UN some countries made an effort, quickely it became a power grab political nightmare that never actually resolved anything.
In my eyes they are. For what it is worth, with the exception of this recent period (Ukraine), both the US and Russia for the most part made pretty good-faith efforts to comply with what their diplomats signed. I'm sure there were undeclared vehicles that were not disclosed, but I imagine those are not that plentiful, and certainly not in the quantity that would make a strategic difference (thousands sitting spare, ready to fly, all it needs is a warhead mounted an hour before launch).
I view arms control as a reduction of the number of variables in play (launch vehicles, warheads, etc). Both sides will always retain their God-given right to unleash nuclear holocaust, but both sides also reason they don't need 10,000+ warheads to do that. They can accomplish that goal with far less nukes than war planning 40 - 50 years ago called for. Partly that is due to accuracy having increased rapidly since gen 1 ICBM. There is a great book about missile guidance and accuracy, it is called "Inventing Accuracy". If you can't find a local copy or can't afford it, the Internet Archive has a PDF you can check out and read. Really insightful as to the "why" we don't require 30,000 nukes (1965) -- we only require ~5000 today and the enhanced accuracy makes them even more potent than prior bomb designs (which had a much larger yield).
