I have seen this sort of thing play out in other companies before so I suspect I have a pretty good idea of what is going on. Somebody has got it in their head that they need to compete in the EDA market so they decided to buy Eagle. That turned out to be a disaster so now this person has a grand idea that they will just start from scratch and build a whole new tool, except they have zero EE experience and are vastly underestimating the amount of work required to make a viable replacement. They will invest a ton of dev work on it and push out some half baked pile of crap "minimum viable product" that makes a lousy first impression and then they will keep throwing resources at it but it will never gain any market share. What it comes down to is the purchase of Eagle was a massive failure, they wanted into the market but completely dropped the ball and existing users fled like rats off a sinking ship. Now somebody is trying to salvage their career by committing to build a whole new tool but there is no way they will catch up to other options. As soon as they made Eagle subscription-only so soon after stating they had no plans to do so I was pretty sure it was doomed.
The only correction or perhaps an alternate possibility is: "Somebody has got it in their head that (...). That turned out to be a disaster, so now he/she was let go and they brought someone else to shut this pipe dream and will just start from scratch (...)"
I have seen the alternate scenario happen many times as well... Everybody else is left to pick up the pieces and turn off the lights on the legacy product.
My reading of the tea leaves is that CadSoft's principals reached the end of their road with EAGLE -- any proper "upgrade" would require a complete rewrite. This left them with a choice: go ahead and spend the engineering resources to do that rewrite, or ... cash out.
They looked at their resources and the amount of time and money they would need to spend on that upgrade. And they looked at their market and asked hard questions, like, "Will our existing user base pay for an upgrade? Will they see value in paying full price for that upgrade, so we can recoup our investment?" Those of you who paid for the full version of EAGLE -- answer that for us! I realize that this is a rhetorical question.
CadSoft also knew that a significant part of their user base only cared about the free version, and those people were never going to actually pay for anything.
And surely they saw Kicad as being a solid competitor, with active development and a growing user base. This must have played a big part in CadSoft's future planning.
Thus, a decision to look for a buyer. Enter AutoDesk, a company with deep pockets and an idea that offering an integrated PCB and mechanical design suite makes sense. (And be honest, gang -- it actually does.)
It would seem that AutoDesk did not know what it was buying, even though I find that hard to believe. Surely AutoDesk's Legal Sméagols told CadSoft to open their books and their code. More likely they underestimated their ability to merge the CadSoft code mess with Fusion. Who knows. At any rate surely the team at AD who made this purchase happen is already gone.
That they say Eagle development is dead and their future is "Fusion PCB" tells me that the CadSoft project of rewriting their tool is actually in process, perhaps even with the same developers, except it's being paid for by AutoDesk.
And we're all saying, "... but ... Kicad." Indeed -- Kicad is pretty darn great, and it's free. So where does "Fusion PCB" fit into this?
Remember that Fusion360 already has a free hobby license*, a free "startup" license, a free educational license, and the "regular" cost for the standard tool is only US$400 per year. This means that the thousands of people who already use Fusion360 for MCAD will get Fusion PCB for free when it finally arrives, and surely many of these users will try it out, simply because it's there. If it doesn't suck, then maybe it will find a user base. I mean, can it be any worse than Altium's attempts at lower-cost PCB software?
And surely the idea that a low-cost PCB package offered by a major software vendor -- that is, one which would hopefully provide support -- appeals to corporate customers who might not trust a free open-source tool like Kicad.
* The support for electronics design in the hobbyist version is limited, but it is not in the startup version.