what do you mean by "good" steam ??
The "good" vapor, the one that "works" has not been characterized by anyone, so that we can try to reproduce it on any vehicle with a good chance of achieving it quickly.
It can not be said with evidence that it is dry steam, or wet steam with such a quantity of water droplets of such diameter, that circulates in such a range of temperature and pressure with such a percentage of saturated air in water....
When you find the "right" parameters, empirically, it works. As soon as you leave it, there is no more effect on consumption. To really know what's going on, it would take fierce instrumentation ... not within the reach of an amateur.
So we hack gently by modifying the architecture of the assembly (GVI or bubbler, positioning of the inlets and outlets of the different fluids), the pipes (diameter length), the volumes of the different "chambers", the temperatures (insulators, size and length of certain components), pressures and flow rates (nozzles, compressors, or venturis), by varying the gas mixtures (air, water, steam, exhaust gases), by modifying the properties of the water (acidity, hardness, additives) ...
We end up arriving at the result more or less quickly and following many other parameters ....
The user manual is still to be written