I thought it necessary to spell it out because your understanding is clearly based on the false assumption that most of the gold investments aren’t backed.
Another poster (tim333) made a great summary of the concept behind GLD (your first link) a few posts up. Those backed assets are otherwise owned and loaned against. IShares is the other one with a similar concept. You can read up on this in the context of Basel 3's net stable funding ratio and the exemption LBMA (the biggest backer of COMEX) achieved.
This is not to say that all ETFs work this way. I believe Xetra is one example of being fully physically backed. But this naturally increases running cost due to storage. Last i checked i believe you pay in the rough region of 0.1% a year in storage cost at the reputable physical gold bullion storage sites as a large customers (tons / room). That sums up. This also explains why physical copper is not something you can easily invest in.
But again, this kind of misses the point, i would really appreciate a definition of paper differing from backed by derivatives. I think you might be onto something here, i ran into this argument a few times now so i believe i might be missing something.
I would also like to reiterate that i do not share the conclusion the initial post made about run away gold prices. I elaborated on that in the other posts.
Sorry for the late edits, had to double check the numbers.
“Paper gold” is meant to refer to a claim on real gold. I.e. if the ETF share exists, there is a small piece of gold backing that and only that share.
A /GC futures contract—-even with physical settlement—- is not paper gold because two parties can create the contract without any actual gold existing anywhere.
I also didnt disagree with derivative overhang being normal. Not sure why you thought it necessary to repeat.