May I ask what the result would be if it did that?

LFOs and Lanes can be set to start over (from t=0, time=zero) on every note. Lanes are a little finicky in triggering them - but you can set them to trigger on ARPs. And your ARPs are tied to clock. Likewise the LFO should see ARP notes and start over if so configured. You probably want the phase to be reset less often than that - but that will do it.

The other is to use first note on - and have an inaudible note as the start of the ARP - with a "forever" gate time (longest possible). This would sync the phase less frequently. Not every note. Until your ARP "lets go" of all overlapping notes.

This is high-level. May be some devils in there depending on what you're trying to do and how you want it to work (in more detail).

May I ask what the result would be if it did that?

Let's suppose you have a sawtooth LFO type motion sequence affecting your filter in a clearly rhythmical way. I would like to be able to set it so that the peak of the first sawtooth wave would be timed exactly to the first beat of an external drum machine or sequencer. At the moment, the phase could be at any point in the motion sequence. Perfectly in sync, but out of phase.

The other is to use first note on - and have an inaudible note as the start of the ARP - with a "forever" gate time (longest possible). This would sync the phase less frequently. Not every note. Until your ARP "lets go" of all overlapping notes.

I have thought of the inaudible note solution too, but I felt that was a bit of a poor workaround really, even if it is rather clever ;). The same with sending an MS trigger CC on start from the sequencer (this works, but it is a few milliseconds late for some reason, so is not very usable).

Interesting idea to use an arp though - I wonder it I can store a MS trig in the first note of an arp and use that. Don’t think that’s possible. I will give it a try.

Just for the record, I did try this.
I could get the new sequencer to send a motion sequencer trigger at the start of a sequence, but it still has that small delay that suggests that the motion sequences aren't tight enough for any serious rhythmical work. This seems a bit daft, as one obvious use for them is to create (for example) a rhythmical filter pattern, tied to the tempo.

I could not store the motion sequence trigger in an arpeggio, sadly.

This is still attempting to work around something that should really be built-in in my opinion (or even just an option).
It's like having a sequencer pattern that stays in sync with the drums, but doesn't care where in the bar it starts - i.e. of limited use.

Too bad motion sequence didn't pick up some of the pattern sequencer features like being able to ritard/advance the playback vs. clock - for example.

That said - depending on the LFO you're using - you may be able to draw a user LFO waveform that has a slightly advanced or delayed "peak" vs. the standard shapes.

BTW: there are "no" poor solutions. Just things that accomplish what you want and things that don't. As a synth programmer, you have to deal with what you have. Sometimes it takes more work to accomplish because there's something "missing" that would make things easier. That's a good topic for suggestion on "ideascale" if you find things that would enhance the user experience. All solutions have consequences - so it's a matter of tradeoffs.

Have you tried the envelope follower?