There's a LOT here. I feel this applies to a lot of decisions.
For instance, if you want to make a product that requires a database and you like building database stuff, do the database stuff last. Do what is difficult first - fail fast.
The easy or default route will always be well known to someone.