I had to read it twice, I don't think any of us can even make up something as absurd. How does any of that solve anything? Castle nut torque value is 38 to 42 ft/lb, but it's irrelevant, except in a broader context, i.e. if he got that value wrong, I'd be skeptical of other things he may say. I'd be somewhat puzzled by the symptoms, I think elimination of variables would be a good start, which none of the actions he mentioned would do. The lower is the first and most obvious piece, start with that, which you have logically concluded as well. I'd return the lower and get one from SOLGW, SIONICS, BCM, Colt, Noveske, etc.
Timing of the carrier is a thing, to be sure, I've only dealt with it when troubleshooting cycling issues in automatic settings where the buffer is too heavy and timing was messed up for the auto sear and release of hammer.