A thought that I have been juggling for a while 'how much is this mod going to hurt my performance?' so it occurred to me, why not have a filing system for that?

think of it this way, every mod gets a number between 1 and 10. 10 being a lag super storm, and 0 being no impact at all, if not an improvement on performance.

If such a system could be developed it could help everyone trying to get here game just right.

There are some pretty big variables to navigate. Primary among them being your system specs - which are almost as many in number as mod users.

But even leaving that aside, you'd have to separate the mods into two categories: CPU-intensive and GPU-intensive.

And some mods would start lagging only when conflicting with other mods on particular records. Too many "unknowns" in the equation to define the index you want, simply saying it is impossible to be objective and will offend mod authors.

I don't think judging the performance impact of one single mod isn't all that useful. It's more about how a whole lot of mods incrementally add up performance costs. Getting the optimal mod setup is all about how you choose to balance things. My rule of thumb is that you cannot run a stable game with both lots of graphics-havy and script-heavy mods.

What I do to balance out my performance is that I don't indiscriminately install all the highest resolution texture mods. For interiors I can go the extra mile with eye candy, but since exteriors are the most resource-intensive environments I generally don't install 2K resolution texture mods. 1K works just fine, and I think things like post-processing effects such as ENB more than weigh out the importance of hi-res textures.

With regards to script mods I am very careful about stacking mods that continuously run script at regular update intervals, since mixing those is the highest contributor towards instability and script lag. You'll have to gauge yourself what kind of mods work that way, since most aren't going to explicitly spell out how they technically work in the description page. A mod like Frostfall which is always checking temperature would be one such mod, and things like Locational Damage or Trade & Barter probably fall into that category as well. You'll have to be very choosy about what you prefer here to avoid overloading the engine (I don't think your CPU makes much of a difference in this case).

I would just be another thing people would argue about. Who would you trust to assign the impact index to a particular mod? A committee?

Don't you mean "is all that useful" since you started the sentence with "I don't think"?

Adding to latency is the main problem when you add a mod. There are some latency measurement mods out there, such as the ones in Convenient Horses and Jaxonz Diagnostics. A person can always check to see if latency increases after a mod in installed, report results on that mod's webpage. More would require some sort of committee, I believe.

Save game file size would be important for mods as well. Another metric affecting performance when loading/saving.

Another would be how many active scripts a mod tends to create. Some create active scripts on a lot of objects over time. So the script load increases over time potentially.

However, if you want such a thing you will have to make such a thing. That is how most mods. addons, tools, etc get created. Somebody felt it was needed so they made it. If your not willing to create it yourself perhaps it is not worth being created?