I had some results for a while. Iâ€™m not sure how relevant they are but for me personally it was helpful. I understand better what happens with the glue. What makes this experiment unreliable is:
1. The torque wrench. I didnâ€™t want to spend more than the $70 I paid for it, which means it might not be 100% consistent. Still, for a range estimate I think it was good enough.
2. Didnâ€™t have a variety of pins and holes dimenssions, so none was a very loose fit.
3. Had only one 3 cp thin glue and two normal ones you get in a store.
I drove in the pins, some new, some old. Numbered them and poured glue, 2, 3, 5 drops at the bottom and noted that on a piece of paper.
After a day I measured torque. All lost a 10-15 % of it. I put some more. Checked in a few days, no improvement.
On another block piece I did the same with same results but I went ahead and applied at the top too. No higher torque regardless of number of drops.
Next, I cut through the holes with the bandsaw.
I can see 4-5 mm discoloration and shinny surface at the top and bottom, so thatâ€™s how much the glue penetrates when pins are a tight fit. Only after applying 2-3 times did the glue pool at the bottom of the pin.
I also cut one old empty hole to compare: no shine or discoloration.
My conclusions (guesses really) are, when a pinblock is tight, applying glue works more like a lubricant, gives a smoother surface and reduces grip. Normally a technician would never treat a piano unless itâ€™s really bad and loose, that's maybe the reason there are no reports on loss of grip. So for a succesful result of a glue treatment, space between pin and wood might be necessary.
Thin or not, the glue goes in a few mm, but not dripping through to the other side, at least not the glues I used. While possible, I donâ€™t think other glue types do that either.
Applying glue both sides vs only at the bottom didnâ€™t show any significant differences, so I wouldnâ€™t flip a grand, too much effort and risk for unsure results.
I think trying to measure torque on a piano is not very useful beyond getting a rough idea of the general tightness of the pins, because the static torque slightly changes with pin position change.
Bottom line, with a bad, loose block I would do it, it might work, if it doesn't, new pin block becomes necessary.