@KristijanL proposed just such a modification a while ago.
But I believe @k_vanlooken has outlined a systematic approach.
The system you have seems to have some intrinsic design flaws/ issues, but it can obviously be run adequately under certain conditions regarding fuel type, condition, size, or it wouldn’t exist as a commercial design. It seems you are trying to operate it outside of the design parameters with regards to fuel.
As Koen states, you are trying to experiment by varying too many operating parameters at once, before knowing the true operating characteristics of your machine, or having experience with the operator needs and intuition.
The rotating stirrer you have installed might be contributing to bridging. Unless it is in very regular motion it probably impedes down flow of chips.
The fuel size is too small and packing too tightly causing bridging and failures of steady downflow and proper air flow for combustion to properly continue.
Maybe just try to run the system on only charcoal, then extrapolate.
Keep it simple. Go back to basics. Get it running in the simplest way, then work from there.
Eventually the nozzles are probably a good idea, but figure out your machine first, then progress.