In the case of the glimmer build - I really think the optimizer is working well right now. I know that if you check other sources of information they will all tell you to stack haste, haste, and more haste… but I really don’t think that is necessarily or even truly optimal from a numbers perspective.
The optimizer does recognize Haste as a good stat and gets a lot of it, but stacking it too much to the exclusion of other stats isn’t really going to get you maximum output, based on the tests I’ve done. We’re the only people using a simulator to test healing builds, fwiw. Other theorycrafters will tell you that you can’t sim healers, but I’d argue that we’ve created a simulation for healers that actually does a very good job of approximating the game enough to get actionable advice. That is generally why you will see our recommendations for stats differ from other sources.
I have updated the simulator to play the glimmer build how people are generally playing it in both the mythic+ and raid healing script, so I feel good about the stat recommendations that are a result of those simulations.
All that being said, haste for healers is a highly subjective stat. Some people just really want a lot of haste to feel more responsive. That’s ok! We created the customize feature for this very reason. You don’t have to worry that using the customize feature will result in a very sub-optimal result. This is why we group the options into “tiers” - anything in Tier 1 will be functionally equivalent in output - you can’t play the game enough to really notice. Anything in Tier 2 will still be very good, but you might be able to notice a difference in output even with the amount of trials a human can do. Tier 3 is viable, but probably what most people would consider “casual” as far as output goes.