It's not the treads, it's the flange.
I decided to fix the remaining NPT/NPS miscogony below the waterline by replacing the head intake line thru hull. (Here, I'll try one of those big pictures.)
I was able to get this smaller thru hull (It turned out to be 3/4 not 1/2 as posted above) out easily and having it, the Groco adapter, and the ball valve all together gave me another chance to look at the thread issue. Here's what I found:
1) The thru hull was surprisingly loose in the adapter. I could make it click until the end was hard up against the bottom. Since it has to be cut a bit short to be sure of fully taking up, this is a joint that will never be tight without tape or compound.
2) The NPS thru hull went fully into the ball valve and bottomed out. It could be wiggled a bit when backed off two turns.
3) The 3/4 NPT spigot on the adapter went only about half of the thread length into the ball valve with about the torque that could be applied in the field. This is only 3 - 4 threads. It could also be wiggled when backed off two turns but not quite as much as the NPS thread.
4) Just to be sure this wasn't a mismatch with my original 1980 thru hull, I took the parts into the store and tried a new Groco thru hull. The results were identical.
Conclusions:
1) The physical strength of the NPS into the NPT is probably greater since nearly twice as many threads are engaged and the end of the male fitting bears on the bottom of the hole.
2) The fluid tightness of the NPS into NPT is certainly less due to more tolerance between the metal surfaces. This might be an issue at water service pressures but at the heads that thru hulls experience should be well within the capability of the pipe dope to seal.
3) These findings are a function of the relatively short thread length of the ball valves. They would not apply to full length tapers of couplings and other fittings.
4) Your NPT ball valve is not going to break off of your NPS thru hull.
BUT, big big but:
This is a frequently used valve. Even though it isn't unusually stiff, I could clearly see the effects of the strain and working of the small thru hull nut on the plywood backing block. A few more years and water getting under the nut would have broken down the wood and the backing block would have started rotting. With just normal valve forces, the strain at this point is clearly unacceptable. NPT ball valves on NPS thru hulls should be replaced for this reason alone.
I decided to fix the remaining NPT/NPS miscogony below the waterline by replacing the head intake line thru hull. (Here, I'll try one of those big pictures.)

I was able to get this smaller thru hull (It turned out to be 3/4 not 1/2 as posted above) out easily and having it, the Groco adapter, and the ball valve all together gave me another chance to look at the thread issue. Here's what I found:
1) The thru hull was surprisingly loose in the adapter. I could make it click until the end was hard up against the bottom. Since it has to be cut a bit short to be sure of fully taking up, this is a joint that will never be tight without tape or compound.
2) The NPS thru hull went fully into the ball valve and bottomed out. It could be wiggled a bit when backed off two turns.
3) The 3/4 NPT spigot on the adapter went only about half of the thread length into the ball valve with about the torque that could be applied in the field. This is only 3 - 4 threads. It could also be wiggled when backed off two turns but not quite as much as the NPS thread.
4) Just to be sure this wasn't a mismatch with my original 1980 thru hull, I took the parts into the store and tried a new Groco thru hull. The results were identical.
Conclusions:
1) The physical strength of the NPS into the NPT is probably greater since nearly twice as many threads are engaged and the end of the male fitting bears on the bottom of the hole.
2) The fluid tightness of the NPS into NPT is certainly less due to more tolerance between the metal surfaces. This might be an issue at water service pressures but at the heads that thru hulls experience should be well within the capability of the pipe dope to seal.
3) These findings are a function of the relatively short thread length of the ball valves. They would not apply to full length tapers of couplings and other fittings.
4) Your NPT ball valve is not going to break off of your NPS thru hull.
BUT, big big but:
This is a frequently used valve. Even though it isn't unusually stiff, I could clearly see the effects of the strain and working of the small thru hull nut on the plywood backing block. A few more years and water getting under the nut would have broken down the wood and the backing block would have started rotting. With just normal valve forces, the strain at this point is clearly unacceptable. NPT ball valves on NPS thru hulls should be replaced for this reason alone.