Great, thanks for that, spiral!
So the relatching works now, in sync and all, that's nice. Unfortunately, the arp is still behaving quite odd, and hard to get predictable results with... The only way to be certain of what will be the first note of the next chord is to press the particular 'first' key of the next chord just after the 'last' key of the previous chord, but you have to make sure to NOT press any other key until after the first pressed key has been triggered, THEN you may (quickly) press the other keys for the chord. I slowed everything down ridiculously slow, but no matter how early I press the first key, if I press any succeeding key after that BEFORE the first key has triggered, then there's no way to know which will be triggered first, no matter if the first key is pressed WAY ahead of the second. I actually found the second key to be triggered more often, but didn't find any way to predict which would trigger first consistently.
So, in effect, when launching the next chord, I have to make sure to press the first key ONLY until it has been triggered, then QUICKLY press the other keys before the first retriggers.
I cannot get any predictable results of the first note triggered in a real, not even a somewhat up to speed, scenario. It stays in sync though, so I suppose it could work for the random arp mode, which I rarely ever use, as I mostly want predictable patterns. So, it's still pretty useless to me personally. If I'm doing a 5- or 7-step pattern for instance, it's essential that I can hear the beat of the root note on the ONE every time, even when switching chords, otherwise it loses it's purpose, the way I use it.
I don't know why most other arps I've used simply locks on to the first key you press, no matter if you press the other keys before the first has been triggered. As long as the particular key is pressed FIRST, it is triggered first. It just works. But not on this machine, unsurprisingly.