Except you're not really solving the right problem. The problem is not that the standard library is bad (It could be shite, it could be gold, it doesn't really

Except you're not really solving the right problem. The problem is not that the standard library is bad (It could be shite, it could be gold, it doesn't really matter). The problem is that regardless of what's in your "standard" library, you will have a VERY long tail of users interacting with your code in a user agent that just doesn't support it. Period. So I think you can rephrase this whole conversation as not "make a better standard library!" but instead as "how do we bootstrap a better standard library into user agents that haven't been updated in 10 years?" In which case suddenly NPM/Babel/Webpack start to make a lot of sense.