I spent some time trying to work out how the Measure actually schedules things, and built a small tool to help. Sharing both in case they're useful.
https://dayn55.github.io/Bard-Rework-Idea-V3/Measure%20Timing%20Lab.html
It has a live timer you can keep on a phone or second screen. Tap when your charm fades and when it lands, and it logs the gap against how many buffs you're running. It also shows an 18 second loop of your Measure songs so you can see what should be cycling in the background, and there's a checklist of tests if anyone wants to help pin this down.
Cheezo, your question is exactly what I ran into, so here's how I believe it's working. This is just my read from testing, not anything official.
Beneficial songs stack in the Measure up to the cap. Detrimental songs can be added, but only one at a time. Every detrimental song removes a beneficial one, regardless of whether you're at the 3 or 4 song cap. Cast a second detrimental song and it replaces the first one and takes another buff with it. So your 3 buffs, then a DoT, then a second DoT, ending with 1 buff and 1 DoT, matches what I'd expect.
The other thing I noticed: mez, charm and DoTs keep renewing on the target they were cast on, regardless of what you currently have targeted. That's really interesting and opens up a lot of possibilities.
I logged 25 charm gaps, timed from the charm fading to the new charm landing:
Charm only: 2.1 to 3.6 seconds
1 buff plus charm: 8 of 10 between 2.1 and 6.5 seconds
3 buffs plus charm: 9 of 12 between 4.3 and 14.8 seconds
That fits the Measure walking through its songs in a fixed loop, one 3 second cast each, with a broken charm waiting for its turn to come back around. More songs means a longer loop and a longer possible wait.
How it felt: I had a lot of fun running 3 buffs with charm on a constant loop, recharming the mob automatically. It made charm kiting very easy. It was also a bit boring, because doing anything else would ruin the setup.
Overall I like it, and it's not that different from what we had. It just has some complexity in how it works that isn't obvious yet. A few things I think would make it easier to understand:
An indicator on the gems showing when a maintained song is wearing off.
Detrimental songs showing up differently from beneficial ones, so you can see at a glance what's running and what's about to get bumped.
Gem slots like cantrips that hold the sustained songs. That would give people more control over what keeps running and what doesn't, instead of the current push-out behaviour deciding for them.
There's more to explore here, and I'd encourage people not to declare this ruined just yet. It takes a moment to get used to.
One thing I didn't get to test: whether a song pushed out of the rotation ends sooner than one you cancel yourself. I'm hoping to check that during the stress test Saturday. If anyone else runs the tool, I'd love to compare numbers.
