At the end of the day the front-end is heavily tied to the visual design of the site. When the visual design changes drastically it makes sense to throw out the tightly coupled logic with it, and so front-end work is more ephemeral.
In this environment there is no incentive to make things last long term. Frameworks can make breaking changes version to version, and websites may change their technology completely in each iteration. It doesn't matter because you will just throw it out in relatively short periods of time.
I'm very interested as to why this phenomenon - not having an authoritative way to vet third party libraries - seems to be a uniquely frontend concern. What have the dozens of other, battle-tested languages done to solve this? How do I know that a Maven package or Rust crate is legitimate, useful, and sans-malware?
It seems like this is a question fundamental to any programming ecosystem, that _must_ have a reasonable solution which balances developer time assessing a potential library with the utility of that library. I'm very interested in some strategies the JS ecosystem could take to foster a culture or platform with well-known, well-supported standard libraries.
The one thing I would add is the JavaScript standard library is deserving of some blame for the micro-library problem. Being bundled with the browser or server side in node doubling the size of the standard library to avoid situations like the left pad library being pulled from npm seems like a no brainer type of decision. ES2017 adds the padEnd and padStart functions is a step in the right direction. There is movement in the right direction but I would like to see the JS standard library at a minimum match the ruby standard library and eliminate a bunch of the micro dependencies that creep into libraries.
Self-promotion is a plague in this industry. First page of Google results regarding any JS framework topic is some marketing folks aiming to be hired/contractors, promoting outsourcing/nearshoring location, without any meaningful content created. Then this laughable community celebrities and "authorities". Crockford, Zakas, Souders, Resig - ever heard of them? In three years nobody will even remember or care about that Angular-Google guy, or React-Facebook guy, or ngrx-whatever guy... except those who end up maintaining some legacy codebase and cursing the universe.
In this environment there is no incentive to make things last long term. Frameworks can make breaking changes version to version, and websites may change their technology completely in each iteration. It doesn't matter because you will just throw it out in relatively short periods of time.