I would argue the opposite actually: when pointing nearly straight down, the compass/heading indicator is much less stable; we should instead try to be as level as possible when lining up the "axis" of the formation, in order to get a good reading on the compass.
[snip].
That was my next effort

-15 Degrees of tilt was the best I could do with a round number (kept bouncing on the planet if I got any lower) and still keep the important bits in shot (the centre mass, the two spires and the experimental control spike I'm using as the back of the "arrow"). I also made sure that rotational tilt was zeroed too this time.
I did notice that almost directly above the barnacles was this...

Must check co-ordinates to see if there are any barnacles on the opposide side of the planet, without a clear view of this part of space.