As long as there are no board schematics provided, things are not reparable. Just certain modules containing 100s-1000s of components can be changed (and thrown directly into the trash).
I recognise that there is environmental and functional value in repairing a $600 part, but the reality is that it’s not currently economical, even with schematics.
With schematics, in certain scenarios, absolutely.
a small shop can do it if they grow slowly. a component-level repair shop run by Microsoft which did not swap parts, but repaired them, can not scale quickly. there just isn't that much talent available to do this, and even when there is, it would be tight, economically.
it costs Microsoft less to build a main board than the price they sell it for. the hourly time of the repair person multiplied by the time to diagnose, repair, test, document, and clean a main board would quickly go beyond what a new one costs Microsoft.
component-level repair just is not feasible at scale.
should they make schematics available anyway? yes.
Another aspect to this is that it board repairs could be made significantly easier if companies bothered designing boards with proper test points and expected values documented in a diagnostics manual. Just take a look at how old school oscilloscopes and other test equipment was designed with not only full schematics but well defined diagnosis procedures.
Doing these programs is just hindering the real right-to-repair cause. They pretend that these programs help, but once they get bored of selling these parts 3 years later, nothing changes. Everything becomes e-waste at the exact moment when it was planned in the original powerpoint they created before starting the development of the product.
At the same time if you know which capacitor to change, you will be able to do it 50 years later too (with a bit of hand-dexterity) - or you can ask someone else to do it for you. It is not expected that whatever basecomponent will go out of production (with the exception of specialized MCUs). An "xbox controller left trigger sensitivity assembly (rev 6)" on the other hand was never meant to be manufactured longer than the life of a domestic hamster.