so this is my point: if a knife owner with a degree of competency and care [which we have established are likely appalled at the second scenario] were to notice a fault in a knife, successfully disassemble said knife, but are UNABLE to right the problem at hand, I see absolutely no reason why a properly reassembled knife cannot be sent back for repairs.
it could be I'm misunderstanding the arguments being presented. I have no expectations of getting a REFUND if I disassemble a knife. I DO however expect not to be denied a REPAIR of a knife simply on the case that I took it apart to see if I could fix it myself and have NOT caused any additional damage to the internals.
that's exceedingly hardline and probably won't sit well with most people if Spyderco reverts to taking up that banner. ie perma-tite et all. but oh well, caveat emptor and such.
I think you're agreeing with most other people in this thread, but just saying it differently.
If you buy a knife, notice it's defective OR decide you don't like it, and immediately send it back to the retailer for a refund per their return policy, that's fine.
If you buy a knife, notice it has issues, take it apart to try to fix but cannot, and then send it back to the
manufacturer for a warranty fix, that's fine too. Likewise, if you take it apart, find issues that you
didn't cause, and send it back to the retailer as defective, that's fine too. To your point, if the manufacturer voided the warranty just because you took it apart, many (most?) people would be upset.
On the other hand, if you buy a knife, take it apart for any reason, put it back together, and return it to the
retailer claiming it's new, unused (i.e., you're not claiming a defect), that's where most people would have problems.
Edited to add: the sticking point here is whether any issues found after disassembly were caused by the user, or were pre-existing conditions. That's probably the biggest risk of fraud for the retailer and manufacturer (I didn't do it! It was like that when I got it!) and I'm not sure there's a good answer to that, other than just assuming a certain degree of fraud will occur, and pricing that into your cost structure.