I agree, up to a point. NU5 increased the agility of protocol development since we no longer have to run a trusted setup to make circuit changes. This means we can now make improvements to the Zcash protocol—and make security fixes—much faster. This benefit shouldn’t be discounted.
That said, in my opinion, NU5, taken on its own, was a net negative—it forced ecosystem developers to re-write a lot of code for Orchard support, along with the other resulting breakage you mentioned, which had significant costs, and it did nothing by itself to improve performance or UX. It’s an incredible technical feat, and a major contribution to humanity’s knowledge and abilities, but it did not directly advance any of Zcash’s goals, at least not considering the amount of time, money, and effort that was invested into it.
In my opinion, the opportunity cost to developing NU5 was unreasonable given the urgency I perceived at the time, and I was among the few people who objected to it, favoring the development of anonymous communication features, which could have solved the scanning/syncing performance issues, as well as—as I recommended in the previous link—focusing on more basic low-hanging-fruit UX improvements, like developing libraries and protocol improvements that would make Zcash competitive with PayPal’s API.
I’ve repeated many times—and never felt heard—that Zcash has invested in deep, technical protocol improvements that have not translated into better usability and more-marketable products. I’ve said many times that the way to do this is to design a product first, prove that it’s marketable by trying things and seeing what works, and then making the deep technical changes to make the product a reality. Our historical strategy of making deep, expensive, technical changes in a hope we’re heading in the direction of a product people want doesn’t work; I’ve learned this lesson over and over from watching friends try to build something first and then market it, it just doesn’t work that way.
My hope for the NU5 retrospective is for the community to learn this lesson. I think if it’s solidly learned, Zcash stands a chance at changing course into a direction that ends up helping the millions (billions?) of people who need private censorship-resistant transactions.