1. Try your hypothesis, see if it actually works. This debate will go on forever.
I will - as soon as the two new PCV valves from Rock Auto arrive.
Thus far, just by playing with a GM PCV valve in my hands I can see how it behaves and, when I blocked off it's connection to the inlet manifold with the engine running I increased the signal to the mixer by some margin on opening the throttle - exactly the aim.
In essence, what I am doing here is akin to setting up a carb (with all that entails) but with the luxury of final (but slower acting) electronic control. The better set up the 'carb' is, the less reliant on the electronic correction correct fuelling is. Basics first, correctives second.
When I'm doing this stuff, I can read live data on a lap top connected to the LPG ECU and as I open the throttle from idle I can watch the mixture go lean and the actuator open to compensate and watch it go to correct mixture. But in that moment between opening the throttle and the above happening, I can 'feel' the hesitation and if provoked a backfire is the result. Who wouldn't look for a solution to that on their own engine?
Keep in mind that crankcase ventilation improves oil and bearing life; you may not have full results of your modification unless you do oil sampling on a regular basis, and the bearings may not show problems for a hundred-thousand miles.
Yes, but remember that there will be little in the way of liquid fuel in the crankcase of this engine and as the oil film at the rings is undiluted with gasoline then I can expect a better seal there also so a cleaner crankcase. I very probably will do sampling as the oil I use is a 'long life' oil and the length of that 'life' is ascertained by oil analysis. There will I think still be enough circulation to remove water vapour. The oil I use has a tackifier which enables it to cling as it cools on shutdown so there's a degree of corrosion protection there.
While I aim to preserve as much circulation as possible, it is worth noting that a turbocharged motor on boost when blow-by will be at its worst dispenses entirely with circulation as the check valve in the line to the manifold must close to prevent boost pressuring the crankcase, and breathing (but without circulation) from its crankcase inlet is all that is then available. That 'breathing' is into the vicinity of the air filter - just as was done for decades before emission legislation mandated otherwise. There is some latitude here in what I plan!
2. Study what propane/LPG forklifts or other industrial engine designers are using to "fix" this problem.
If I can get near one.... Covid really has made access I once enjoyed a whole lot more difficult. Generally, here in the UK, indoor forklifts are electric.
Standard practice for converted cars is as I have already outlined (and am trying to avoid). Worth mentioning that injection systems are becoming the norm for LPG conversions and they neatly side step the issue.
3. Contact the folks who made your gas mixing valve, see what they recommend.
This route I am investigating is on the advice of the company who sold me my new reducer/vaporiser. Their view is that I either accept routing LPG thorough the crankcase or devise an alternative arrangement as undoubtedly I am currently losing vacuum signal to the mixer on opening the throttle.
On an informed LPG forum they suggest the mixer being too big (recognising the lack of signal on opening the throttle) so while there is confirmation of that loss of signal, there is no explanation as to how a venturi
loses sensitivity with increased air velocity through it. It pulls the LPG at idle adequately then loses the signal when the throttle is opened. Venturis don't do that - they just don't!
New PCV valves will be here in a couple of weeks - then we'll know. Cheers!