It’s possible you don’t have experience working in a large company? It’s a bit of a miracle that any hardware project gets done in my opinion. Delivering on time is enormously difficult. Adding “would be good“ features like maintainability means cutting something else, likely more cost, and a schedule hit. That costs the whole team at review time.

Even the most devoted person can have trouble moving the needle without exquisite political sensibilities. There are plenty of people working in toward the goals you describe inside Microsoft already, but it’s insanely hard to get things like this to critical mass.

Very often in real life we have only bad choices.