This is not really achievable and would not really benefit mod developers, nor players, nor FAF base game developers.
I understand your point and it is a shame to see mods break due to these changes. Still, I don't think there is an actual solution to this problem in any way.
Let's assume we introduce a new featured mod "stable" (better name would be legacy as it will never be able to change again).
Splitting the ecosystem
Your suggestion implies that we have mods that would run against this stable version. They will not be able to run against the latest FAF version. So eventually we will split the mod ecosystem into 2. Apart from the implications that we would need to manage mod metadata which version works where, it would get confusing from user experience:
Is a mod working for stable or latest? How do I install one or the other? How do I delete or the other? How do I activate one or the other in a game? How do I rate or comment one or the other version?
As a mod uploader you also need to consider the same on uploading of mods.
Let's assume we overcome this. The next problem is
Version conflict with GPGnet protocol
Please note that changes to the GPGnet protocol have been done. These are required for further having working game results
Thus, no older version would work flawless with FAF. Somewhere the result parsing or some other aspect of the game would break with regards to the FAF backend.
Also introducing a stable version would mean that it must never be touched again or backported to stable at all costs.
I don't think the game team will ever agree to this, as this causes an unforseeable amount of complexity and work in the future. We have barely enough manpower to test current features, no we'd need infrastructure and manpower to test another stable branch.
The damage is already done
Let's assume we pick any* version as the main version.
All current mods would be broken again, as they don't work. All mods created before that version are most likely broken too.
So only continuously updated mods of a certain age have a chance to find a working upload for that version. But: It's highly likely that these mod authors already arranged with the process of continuous uploads...
(* also must be any with the recent gpgnet protocol)
Also if you consider going back to the Steam version there are other places where mods are uploaded already.
Alternative approaches
A stable API
Why can't FAF just offer a stable API like other games?
Writing a good API is really, really difficult ans has many disadvantages. There is an open source project with hundreds of full time contributors that has a similar problem: The Linux kernel. You can read up there, why the don't support a stable API and I think it can be mapped to FAF almost 1:1.
TL;DR: Not happening
Compatibility version tagging
Why don't we just tag mod (versions) with compatibility metadata and pick a matching FAF version?
Ok. That is a first step. Let's ignore the technical challenges and assume we have a list of compatible FAF version to each mod version.
Now what? You launch a game and select a mod - woosh to late! - the mod must be downloaded in the right version before you launch the game. So effectively this means mod selection moves into the client. Huge tasks, many benefits, let's assume we did it.
Hmm now depending on the mod you get a "random" FAF version with random balancing.
Now what happens if you select another mod? Sure. Okay soo maybe there are some overlapping supported versions, but now you play potentially some old versions of the mods.
And again now you run into the problem of a potential outdated GPGnet protocol version.
TL;DR: Huge effort, little benefits.
However! Compatibility tagging would at least help users not to play broken versions and get disappointed. It might also be helpful notifying mod developers that something broke.
Incorporating fafdevelop
We already have a "stable" and a "development" branch.
faf
and fafdevelop
.
There is currently no way to publish and test mods against fafdevelop
only under the same mod. Except from doing this alone on your PC which sucks ofc. Not sure if tagging mod versions for fafdevelop would help the development flow.
Eventually this might help interested mod developers to keep pace. But this does not solve the original issue of mods breaking after the developer stops keeping up.
Summary
As said before I see no solution to this problem except for stopping development to FAF game forever. Or mod developers just publish for Steam again and not use FAF.
Some mods follow a dual publish model for latest FAF and vanilla like e.g. BlackOps.