I've said this before, but it's worth saying again:
Define "improvement".
Because the best UI (as in unquestionably the absolute best) is the UI that you're already an expert with.
Can there be better UIs later? Yes.
Can there be better UIs for new customers? Yes.
Can there be a better UI than the one you're already an expert in? NO! At least not until you do the work to become an expert in the new UI all over again.
At a minimum, you're asking users to trust you that re-learning their UI will ultimately be a better experience for them. Because it will absolutely suck for your existing user base when it first changes (this is why UI changes are almost exclusively met with negative backlash from the current users).
Can you do that? Sure. Do companies do it often? Nope. The vast majority of the time, the UI is changing for one of two reasons:
1. The company believes the new user experience is better with a different UI. They are actively trading existing customer satisfaction to improve metrics around new users.
2. The company believes they can drive users to new features with a new UI. They are actively trading existing customer satisfaction to improve metrics around new features.
Neither of those are compelling reasons for a new UI from the perspective of an existing customer.
Broadly - they might still be an acceptable trade for a company on the whole. But don't expect happiness from your current users, and be very wary of the churn and brand loyalty you're burning by making the change.
Don't believe that was worth saying twice quite frankly you're the only one trying to redefine "improvement".
Here's a very simple non-UI example for you. You have learned to cut an onion one way. It takes 5 minutes to cut it that way. Someone else comes along and tells you that you can cut it this other way, and now the process only takes 30 seconds. Is there some learning here? Yes. Are you suddenly incapable of cutting onions or are no longer an expert in cooking just because you learned a faster way? Absolutely not. But overall it is an improvement including any learning process.
Existing customers who remain stagnant because they are terrified of any change are customers who eventually leave when they're marketed to by someone else. It's truly that simple.
Don't get it wrong, this isn't an argument in favor of mindless UI changes (and frankly Thunderbird's changes are garbage) as bad UI updates do exist and should be argued against, but this idea that improvement doesn't matter is so beyond absurd!
