Yeah, this failed for me, too. And I wasn't even using it - I just got in the car one day and noticed it was gone. I suspect the button was bumped or caught on a backpack or purse getting in or out of the car, because the part was no where to be found.
This is a lesson in product management: if you implement a completely new, "cool" design, expect it to fail in new, "cool" ways. In this case, the failure was not just that the damn thing is prone to breakage, but that a very high fraction of the user community finds it silly, or outright hates it. This is a clearly self-inflicted injury.
This button and wheel design was clearly driven by aesthetic factors and a desire to have a market-differentiating steering wheel design. But when you already have a huge amount of design risk with the critical infrastructure of the car, why-o-why would you take on additional risk for something that is not part of the core product requirements?
If I am not mistaken, Tesla's first Model S steering wheels were identical to those on a Mercedes. That makes sense. By purchasing a proven subsystem, you benefit from millions of user hours of "testing" and de-risk your decision. The steering wheel (and associated stalks) is a high-touch piece of the UI, with lots of functionality packed into it. If you get it wrong, you upset your customers in a very "in your face" sort of way.