You can start by checking the fueltrims to find out if the mixture is lean or rich.

Did you looked at the mechanical side? Especially with an engine running on lpg you can have a compression / valve problem.
If the basics look good then you can check the ignition.

Click on the channel number above the waveform. There you can activate the filter.

After the voltage peak you see the closing in the voltage trace. Very interesting. So the voltage trace somewhere shows the actual physical position of the pintle, sometime after the inductive voltage spike from voltage being shut off? It would show that the pintle didn't close if a piece of trash ...

Use a filter to get a cleaner signal. Lambda signals can be very noisy.

The current ends directly when the ecu shuts of the injector. You can only see the opening in the current. The voltage contineus after the ecu shut the injector off. After the voltage peak you see the closing in the voltage trace.

I don't have any experience with that particular car, but I would check alternator output and power/ground of the ecu. About the pintle movement of the injector, in the current you should see the opening, in the voltage you see the closing of the injector. But in my experience not all injectors give...

Hello Steve, Very impressive the capabilities of the pressure tranducer, and good explanation in your case! I do have a question/suggestion about the degree markers. Is it right that the software uses the 720 degrees markers to know the time of a full 4 stroke cyclus, and the markers used to measure...

This is the capture I was talking about.
I was suprised that a signal like this whas caused by incorrect valve clearance.

Not shure if this is usefull, but I have seen something like this on another car.

I have a capture of an o2 sensor signal that looks like this, with the fast and short peaks in a slow signal. The problem of that car was too small valve clearance.
I post the capture asap.