Apologies if my question was not understood. When you build with npm, and components get packaged in dist, that is the end result of a compilation step. When you use dependencies from npm (or pnpm or yarn), they get installed into node_modules.
When you make changes to code, you have to be mindful of your deploy and build process because if you aren't you may find code you changed that worked is suddenly gone.
So in the case of NPM dependencies in node_modules: If you make updates here, those will be potentially lost the next time that module is updated from NPM. (Like say for a security update)
In the case of any /dist* folder: since those are .js files which are produced by npm build commands, making changes there may work when testing, it may work after copying the dist folder to production, but the moment you rebuild something from the your-extensions/widgets folder, it may wipe out or remove any and all changes made in the dist folder. This is because dist is a folder for publishing, not development.
I feel like this is something that nodejs and react docs don't spend a lot of time on, because publishing/delivery is so near the end of the life of many projects, I'm not surprised that anyone new to Experience builder potentially might also get confused. (I'm not sure how much you have used EXB btw)
So when I asked this question, I was trying to understand, what purpose editing something already built and ready for deployment in a publish folder really accomplishes long term.
I think you can use @JeffreyThompson2 suggestion of copying the widget, of course the practice of doing this may also mean that updates to those widgets by Esri in the future will not be 'cross polinated' into your new customized version. (Many understand this as part of the nature of how custom widgets work, but since it may not be obvious to everyone, this is why I mention this.) It is an approach that may work. (although some WAB widgets are not quite like their EXB replacements)
