I can confirm there is an LFO sync bug. My test scenario wasn't the same as yours but it points to the same or a similar cause:
In my test patch, LFO 1 square runs at 1/4 the rate of LFO 2 downwards ramp, with both assigned to VCO 1 pitch, making a 4/4 pattern of piu-piu-pio-pio which makes it easy to hear the sync or lack thereof. After a period of inactivity LFO 1 stops.
When LFO 1 has stopped, it can be made to run again by setting its rate to the same rate as LFO 2 or higher, but it will not run at the lower rate, i.e. setting the rate lower causes it to stop again. While twisting the LFO 1 rate knob, through the range of lower rates, it sort of start up as you twist the knob, but stops again as soon as you let go of the knob (after one period or less(?)). Also switching the arp or sequencer on and off makes LFO 1 run again.
The same happens on role reversal of LFO 1 and LFO 2, with LFO 2 stopping after a period of inactivity. I.o.w. the LFO with the lower rate stops after a period of inactivity.
I also hit upon an intermittent phase issue for LFO 1 while testing, but that's hard to reproduce consistently.
Ticket submitted.