Sure, he's not concerned about making Wall Street happy, that's well known. But he still must, every single quarter.
I'm confused as to why you think it's a myth? Public companies have a fiduciary responsibility to maximize shareholder value. If any company places some other mission above that, they'll be pummeled to death in the public markets. That but doesn't preclude other secondary environmental or humanitarian missions, as long as they are secondary.
MBAs certainly push the notion that increasing share price on a quarter by quarter basis is the sole obligation of a company.
One can observe from Enron or Purdue or Boeing that maximizing QonQ returns does not always work out. Amazon famously told share holders to bide their time for a decade before turning on the money spigot; turning it on to maximum (for amazon) may end up harming their brand equity in the end if consumers buy counterfeit children's aspirin on amazon and their kids turn purple.
The list of companies that make decisions on the basis of "what would a sociopath do to maximize money before EOQ while strictly staying within what my legal department says is legal" that end up evaporating in a cloud of fraud and tragedy and greed is extensive.
As far as Tesla goes -- if they're in a place where they need to convert owners of S/X to upgrade in order to stay solvent, tesla is 100% screwed. Tesla's market propositions are "we need to convert cars to electric"; "electric cars last a long time and are a good investment even if the upfront costs are larger than an ICE car"; "tesla cars are upgraded even after we sell them because we're awesome"; "there is way more demand than there is production capacity"
If you assume those statements are correct, telling customers "pound sand buy a new car" may earn you a small number of new car sales but will also make it clear that several of your core values are lies.
On the other hand, making a 3rd version of the MCU that has plugs for old (MCU1) and new(MCU2) and newer (plaid) platforms may cost you money to develop the new board, may cost you some marginal number of sales of people who want to upgrade / repair MCU rather than throw the car out and buy another one (which may not be a tesla, if the owner is selling out of disappointment), and some cost in making the replacement part more expensive.
But making / selling an MCU2.5 puts the chit down that tesla stands behind their products and that they're good investments (for the consumer).
A universal intel MCU is a bad "MBA/sociopath" move, but it is great marketing probably not actually an enormous amount of money.
Burning your high-touch customers with yellow screens, broken door handles, dead MCUs, etc where you're constantly giving them $500-$3500 bills in things that are clearly design defects will eventually soil the Tesla brand, and that dirt will affect their ability to sell into other markets. "You want me to spend millions on trucks when you can't even make a damned door handle or a computer that works past the end of the lease?"
We shall see. Tesla's record of "MBA" vs "good company" isn't actually very good, if the OG roadster is any indication.
