The story as I've come to know it goes like this: VW thought their best shot at getting a product to market was to use a collection of existing technologies -- the various modules and code that run the car -- and wrap them in to this new MEB platform. It was supposed to save time so they could beat Ford, Hyundai, Volvo, etc. who were all working on $40k EVs.
They nailed the design timeline, the retrofitting of the assembly plant, battery suppliers, manufacturing... but ran into roadblocks with the software. The ID.3 launch was delayed months awaiting functioning software. The ID.4 was launched in Europe with buggy 2.0 software, and the US launch saw cars sitting at port awaiting 2.1.
VW's problem wasn't just integration, i.e. getting existing tech ICAS to talk to new-tech BMS modules. It was also coming up against intellectual property concerns. Mid stream, VW decided the best way to proceed for the future of the company was fold much of their contracted software into an in-house development. Their software arm, Cariad, formed small offshoot companies with their main supplies, Continental and Bosch, to take the existing code and develop into a VW-controlled product.
The Continental venture took place early on, I recall around the time of the ID.3 launch. But the Bosch deal appeared to happen later, and according to I believe Diess on stage at a event in the last month or two, he spoke as if they had only just recently secured the remainder of the IP they needed to make this transition. So this was more than just will -- it was contracts and lawyers and money and negotiations.
Meanwhile, owning somebody else's software isn't enough, a team is needed to understand and transition it, and VW's been hiring software engineers like crazy to make software an equal 3rd pillar of their organization. I can only imagine that these thousands of employees aren't all focused on adapting old code to the MEB platform. VW is looking towards launching their successor SSP platform in the next four years, and that one will be more of a ground up venture.
It would be nice is MEB was a ground-up product, but it's not, and it would have been at least two years delayed had they started software from scratch. They can't just nuke what's out there with over 200,000 MEB cars already on the roads, and I doubt VW will never take the Tesla approach of just pushing rushed code to their cars.
Whatever they're up against now must be serious. I expected 3.0 to happen with the model year change, which itself was uncharacteristically pushed back to the final weeks of December. No 2022s have been delivered to the US yet that we know of, and the ones in Europe are still running 2.3 (their latest coder; the US is still 2.1). 3.0 must have had a dealbreaker in it because now we're waiting for 3.1, which sounded final two months ago, but has yet to show itself. Heck, even the ID. BUZZ demo fleet was running either on 2.3/2.4, or on a preview version of 3.0 that was kept hidden from the cameras recording the test drives.
So take this as me making excuses for VW if you will, but I'm convinced they are really in a tough patch at the moment, of their making of course but nevertheless in a place they'd rather not be, and they desperately need to get 3dotwhatever software rolled out so they can get this MEB platform and the promised OTA updated on track. They're going to be reliant on this platform for the next four years, so they need to get us all squared away if they want to compete and sell cars. I have no doubt they're working hard, if not for our sake, for their own.