@Kent Dawson yeah honestly 4-6m under moderate to dense canopy is pretty much what physics gives you regardless of the receiver — tree cover causes multipath and signal attenuation that no amount of tuning fully overcomes, so you're not fighting a config problem, you're fighting geometry.
That said, there's a few things worth trying. The 18Hz fix rate is actually your friend here — if you're logging position data rather than needing a single-point snap, averaging over a short dwell period (even 2-3 seconds stationary) at each survey point will tighten that scatter significantly. Multipath noise is somewhat random so temporal averaging helps more than people expect. Also worth checking whether you have multi-constellation enabled in your config — if you're only running GPS and not pulling in GLONASS or Galileo, you're leaving satellites on the table, and under canopy every additional bird in view matters a lot for DOP.
What does your fix quality look like in terms of satellite count and HDOP when you're seeing the worst scatter? That'll tell you whether it's a signal availability problem or a multipath problem, and the fix from there is different.