I shouldn't be doing this!

It does indeed do that when you are printing a flat surface but not when you are printing a cylinder in layers.
You are relying on how well each individual bead sticks to the previous layer.
The problem is the surface is cold by the time the next bead is laid down. At a guess I would say it has no more than 50% of the strength compared to the tensile strength along the line of the beads.

This is how the its maiden finished on the 14th Dec.

The calm weather has spurred me on to do something about it.
As I suspected the motor mount broke.

Entirely my own fault just poor design not taking the stresses properly into account.
So easy enough to redesign it progressively thickening the mount toward the motor and reprinting it.
Using POR's natural "rubberyness" it was not too difficult to cut out the old mount and insert the new one requiring only a relatively small fuselage skin cut out.

A bit of local reinforcing and it should be a good a new!

All back together.

It also has a 9x3.8 rather than a 9x6.
No real reason other than the Cessna Skymaster managed very nicely using a pair 9x3.8 and the Cessna Bits has an even lower wing loading!
Weather looks good for tomorrow.

Another grey raw day but it does fly all right.
Thrown around a bit by even the modest turbulence low down (very low wing loading) but nice enough in the smooth air.
Now I know it flies I will video the next flight just to complete the story.