Folks,
If we were to migrate to a JS framework every time a new one showed up on JavaScript, we would probably be on our fifth or sixth. Gulp/Brunch, Webpack, Parcel, etc... The issue with changing those is that it causes an incredible amount of churn. There are a bunch of articles on how to integrate Tailwind, Bootstrap, etc to Phoenix and they will all become invalid - and this information is valuable exactly for the folks who really don't care about Snowpack vs Webpack and just want to get that part of their stack done. For the folks who do care about this, then you can pass --no-webpack and bring your own.
Of course I don't rule out moving to Snowpack in the future but moving to it so early means an increased chance of churn. Churn now and potentially 1 year down the road if everyone moves to something else. The fact Snowpack is already on version 3,0 means it is still moving fast and that brings even more complexities.
So I don't see this change happening for now... if you like Snowpack, go for it and write how to do it! but definitely premature for a general migration.