Well, Don, that's quite an interesting study.
I note that they only used #6 for the bonding conductor. Considering the nature of lightning - lots of volts, very few amps - I expect even a much smaller wire would be just as effective.
The study often mentions that "shorter seems to be better," as regards the length of the bonding conductor. I find this to be slightly misleading, and I'll explain my reasoning in a moment.
Finally, the study notes that other sources of transient voltages are outside the scope of the study.
Now, for my take:
I have no objection to using a #6 conductor. There are other considerations, apart from current carrying capacity. For example, the need to stand up to mechanical damage, especially from the gardener.
Since we are assuming the electrical source to be outside the dwelling, "nearest" means to me that the best point is at the gas service, at some point between the gas meter and where the pipe enters the house. In my understanding, routing the conductor inside the house to the point where the CSST begins would be 'farther' from the source of the current.
Yet, were one to want to add some CSST at a future date, I see no issues with running the bond wire to and convenient panel - and sizing the bond wire smaller, if the EGC to that panel is also smaller. Perhaps we want code language to tell us to size the bond wire to the EGC of the supply, but in no case need it be larger than #6 (language like we use for the GEC elsewhere).
Which brings us to addressing other causes of these pinholes in CSST. Let's face it, we are getting some mighty high transients from various items inside the house. Electronics, frequency drives, induction heating - all these things that were once only of concern to industry are now showing up in our homes. These transients are being backfed into the power grid- a matter of concern to PoCos. I can see where bonding the CSST at the service end will serve to 'drain' any transients that might otherwise build up a charge on the CSST.
Finally, I wonder if the CSST problems are greater in those areas that also have "stray voltage" issues. That is, could the pinholes be caused by obsolete 'ground neutral return' power grids?
