One could argue that Microsoft's "no game sharing" decision at the beginning of Xbox One, right off the heals of a generation that experienced both high success, but also extremely high costs (RROD recalls, estimates say it cost them over $1.1B), was a bigger one. It seems unnaturally likely to me that single decision by Mattrick's Xbox leadership is the largest component to Xbox One's relatively lower success (58M sales, versus 102M PS4 and 84M XB360), and has cascaded to, possibly permanently, suppress Xbox's potential (Series is at 21M to PS5's 41M).
I tend to think that decision was worse, because it was Just Bad. There was no reason they had to do it. It had very little prior art (digital games, naturally, always did this, so non-zero). It was a pure, absolute failure to read the room, and the XB1 era had so many decisions like this that it seriously paints the picture that Xbox is alive today literally only because of Spencer and despite Microsoft's leadership at the time (pre-release, Xbox One also required Kinect to be plugged in. This was walked back, but it still shipped in the box, lowering margins and depressing sales volume for no reason.)
In comparison: Unity has been in trouble for a long time. They've never posted a profit, at least since going public, and have instead accumulated an aggregate of over $2B in losses. Their main competitor (Epic) is a technological savant of a company, which has a literal money-printing machine in their closet to finance engine development. They never really broke into AAA game production, partly because Unreal is just better at hyper-scale development, and partly because AAA game shops are surprisingly welcoming to in-house engine development.
In that framing, this outcome seems inevitable. There's a reason they made this decision; a very real, dollars and cents, reason, and backing away from it (especially after the outcry, but even without it) isn't going to save them. The best, but unlikely, outcome for Unity and their customers is deep layoffs, focusing hard on ease of use instead of technological prowess, spending years rebuilding the trust they lost, and depressing their stock price enough that they can find a path to going private without losing control. The much more realistic outcome is: deep layoffs, coasting the product, monetizing every addon they can outside of the toxic "core pricing structure", and private equity acquisition.