AtomRC Swordfish - Crash, One Motor Failed Midflight?

The auto launch appeared to work perfectly, turning back to home seemed pretty normal.
As I was passing over the building I noticed a drop in speed and the plane being tugged to the left. I assumed it was just a gust of wind and pushed more throttle to keep the plane from stalling.
Coming down along the tree line, I felt more loss of speed and I was being tugged towards the busy street traffic. I frantically applied even more throttle attempting to regain control and this put the plane into a chaotic spiral into the ground.

In hindsight, I'm assuming I had lost the left engine. If I understood what the problem was at the moment it happened, I may have been able to cut power and glide the plane down safely. Applying more throttle to the right engine only made things way worse.

Total damages were surprisingly not terrible. The foam nose of the plane tore off, and the 2 axis pan/tilt gimbal I had just finished installing on the nose cover was torn off damaging both of the servos. The DJI O3 camera was thankfully OK.

Thanks for watching!

[daniel.pendzich]

UPDATE:
I was able to do some Swordfish bench testing since this crash and I can now easily reproduce the motor stall.

One or both of the motors can fail to start up again after going zero throttle. This is even more likely to happen when dropping to zero and then sharply pushing throttle back up again. at 00:33 I quickly push the throttle up from 0% to 25%, I believe this is when the left motor fails to start up again.

I thought the problem was the ESC braking being on, or the DSHOT600 protocol, but changing these things made no difference, I could still stall the motors on the bench.

I suspected the Bluejay firmware may have had something to do with it, so I flashed the ESCs back to the latest version of BLHeli_S with ESC braking still on and everything else default, and now the motors are working reliably.