I took a closer look at Nature's Balance on the beta. Sunfire is being extended on cast success, not on spell hit, just FYI.
Checked again and you're right, sorry for the wrong info and thanks for the correction - guess I didn't account for the higher latency from Europe and got a wrong impression that way.
In my update (which we will post today sometime) I just removed the armor bonus from moonkin. There really isn't much point in modeling it in the simulator. Also, we don't have the movement stuff in yet. When we get that working we will add the cast while moving effects into the simulator. Incarnation was actually working fine: it applies a 35% damage buff if you look at that spell. The damage buff on the moonkin form spell wasn't doing anything. Someone put that in wrong because they didn't understand how the buff caching works in the simulator.
Removing the armor bonus obviously does make sense, thought I mention it after not reporting some obvious things on several projects in the past and then seeing those not beeing fixed for a long time (just ask Kihra how long "Klaxxi" was mispelled on Warcraftlogs...).
Suspected as much for Moonkin Form, Incarnation and movement spells but same reason as above - better report something obvious which is not mentioned in the bug list before it doesn't get fixed.
Filtering to astral spell school won't work for moonkin form damage buff. It will not buff pure arcane or pure nature spells. You can see how I did it in the update later. I used two buffs, one arcane and one nature, and then blacklisted astral spells from one of them so they wouldn't double dip. We don't have a way to target two spell schools with one buff right now.
Yeah, that confused me quite a bit but seeing as it had been implemented that way, I assumed you added that functionality as well. Will definitely check for similar things in the future and not assume it'll work just because someone else thought it would.
But to get to more serious things:
After that high praise for Feral stuff in my previous post, I've found a pretty big bug with Feral Dot extensions almost instantly after submitting my previous post. Did some more testing and think I've been able to confirm it:
Currently you're using "DotNextTickSec(Rip) + DotDurationSec(Rip)" when extending the Duration of Rip using Ferocious Bite when in Execute Range or if Sabertooth is specced. As far as I can tell, this just adds the time until the next tick and default Rip duration on top of the current duration instead of resetting it to the maximum duration. Using Sabertooth I'm getting Rip durations upwards of 200 sec at the end of a 450sec fight and which also causes ridiculously long durations of Ashamane's Bite (i.e. Ashamane's Rip), as that just copies Rip's duration.
Should be pretty easy to fix by just subtracting the remaining duration of Rip when extending it.
While experimenting with that I've also noticed that DotNextTickSec() seems to always return 0 regardless of which (definitely active Rake or Rip) Dot im testing, so the function above won't even work correctly.
I'll continue taking apart every little thing for Balance and Feral to be able to provide more and more feedback.