And of course if the threads are near the strip point, it is possible that the threads bend a little in operation. When this happens the bolt thread gets a "Christmas tree" look. Not just a theory - we have seen it.
When the thread bends, the clamp relaxes a bit. That lets the case halves move relative to each other and they fret. If they have fretted enough, then a subsequent reef-up will nip the crank. Again not just a theory - we have seen it. Replaced the through bolts, torqued to 30, nipped the crank. Engine no turn and all that - a 700 hour 3300.
The other thing that happens if the clamp releases is that the pots move round a bit. You can tell this is happening if there is oil weeping round the cylinder base. If this happens, the through bolts are bent back and forth and will eventually snap at the pot/crankcase boundary. The cause is bending-back-and-forth-fatigue. If you look round the net a bit on the failure of bolted clamp joints that's more or less the universal cause.
As to the use of Loctite- I am a but confused - Lycoming and Conti say that through-bolt nuts should be lubricated (engine oil is suggested) before they are torqued up .... go figure ... what would they know?