Nobody can dispute making everything in megafactories was too easy, linear, and out of balance. But there are so many issues with your approach to fix it.
PROBLEM 1) You are putting yourselves in quite a position with treating the game as a beta and non-beta whenever it works. On one hand, with the promise of keeping our wealth into the release, we're expected to treat the game as it's an actual launch to compete with other players and organizations. But then there is no support, mechanics can change dramatically, etc. You're also putting at least a 6-months gap between those who could exploit the unbalanced mechanics early and accumulate wealth, and those who couldn't.
SOLUTION: I don't really have a solution to this. Perhaps the best way to operate right now is to announce a full wipe at the end of beta if you can financially handle it.
PROBLEM 2) This update leaked to lots of people for several weeks. The knowledge put those people further ahead.
SOLUTION: Either enforce your NDA, or do not disclose (economically sensitive) details of how you'll change things.
PROBLEM 3) Completely unhelpful talents just to unlock elements are a bad idea. Eve Online learned this lesson over the years and they are doing away with artificial prerequisites to start playing the game. Industry is one of the rare domains where talents actually GREATLY matter. Most production will eventually become unviable overtime without respective talents. It's bad analysis if you thought the lack of talents was an issue.
SOLUTION: If you want to wall element usage behind talents, you should do so behind the existing talents. This way at least people get a benefit alongside element access. A sense of actual progression and no waste of days of training.
PROBLEM 4) I am quite worried whether the game actually has enough quanta supply recipes for sufficient production.
SOLUTION: The game needs isk faucets besides the honeycomb NPC order businesses. I have low-cost proposals for that in the idea box.
PROBLEM 5) Machine-based recipes create lots of issues. It'll be quite tedious to teach the recipes to every single element and to keep tabs on them even in non-megafactories. Moreover, how will our investments be saved? I hope (but am not certain) the recipes will be saved under "dynamic properties" of a machine. But then we won't be able to deploy the same factory from a blueprint. It'll be impossible to move factories. So you're not only investing in the machine, but also in the location of the factory. That'll have even greater repercussions with territory warfare.
SOLUTION: You should have at least made the recipes character-based. Given that you don't tolerate account sharing and ban those who do, this would still achieve specialization, yet make it way less tedious.
PROBLEM 6) Cost is an artificial and unsustainable way to motivate people to specialize. Once people accumulate enough capital, megafactories will begin popping again. Then those who can afford them will be miles ahead of others.
SOLUTION: To achieve truly sustainable differentiation and specialization, there needs to be differentiation in the ways the source materials are acquired. Ore is equally available to every individual. It should take organizational level effort to access some building materials; it should take outside-the-box intelligently developed systems to access others. Moreover, it should make more sense to make one product somewhere and to make another elsewhere (geographically).
