Humans operate by doing each other favours. You'll find you intuitively know who you've done a favour for and who you owe a favour to. It's inherent to being humans and three year olds can do it - as science has discovered.(https://www.en.uni-muenchen.de/news/newsarchiv/2016/paulus_s...) We all walk around with a favour ledger in our heads.
Essentially "here's a chicken, you owe my a chicken's worth of something sometime".
As groups get larger you end up taking tokens as an IOU aide-memoir. And to nail those who try to cheat.
That then evolves into money. Token money. Tokens representing promises.
But intuitive tracking of who owes you favors and who you owe favors to does not scale. The double coincidence of wants problem is not a myth; it just doesn't come into play until you start dealing economically with groups that are too large for intuitive tracking of favors to work--roughly, groups larger than the typical hunter-gatherer tribe.
As we see in the comments above, the gift economy scaled to include Bronze-Age Mesopotamian international trade. It's true that a Hittite in Hatti wouldn't know everyone in Nineveh, or indeed anyone, but he knew the Hittite king, who knew the king in Nineveh.
Or if you doubt that you will ever meet that trading partner again, or if you just don’t trust them and there is no legal framework for recourse if they screw you over...
The 'double coincidence of wants' problem of barter exists. What is a myth is the assumption that any culture used barter as a primary means of exchange. Barter was used on the edges of tribes, clans, etc. between groups of people who do not share the kind of trust that exists between members of groups, who will likely live out the entirety of their lives around each other.
That does not take away from the fact that the double coincidence of wants is not a myth. A lot of card trades depended on both party having a card that the other party wanted when no monetary medium was used. I am not saying that this was always the case but it happened often.
The realization that there is (a) a double coincidence of wants and (b) this can be solved with some sort of currency are absolutely questions already answered if not intuitively than explicitly by money in modern society, as are related concepts like debt and interest. When someone talks about the double coincidence of wants being a myth, they are not referring to the obvious fact that this is a problem money solves, but that this reason, alone or primarilty, is the reason money, as a world-historical concept, came to be in all societies.
So, no, part of the point is it didn't happen. That is, neither direct nor indirect barter. In a non-alienated society, you have many tangible and intangible obligations towards your village-mates or tribe-mates, which include distribution what you produce or gather mostly among others (since what will you do with it personally after all). And there isn't a discrete "chicken's worth" which you "owe" someone in particular. and so on.
That's not incompatible with the existence of coincidence of wants being a problem. Grubers point is that debt is historically the primary solution to the coincidence if wants, not money.
Humans operate by doing each other favours. You'll find you intuitively know who you've done a favour for and who you owe a favour to. It's inherent to being humans and three year olds can do it - as science has discovered.(https://www.en.uni-muenchen.de/news/newsarchiv/2016/paulus_s...) We all walk around with a favour ledger in our heads.
Essentially "here's a chicken, you owe my a chicken's worth of something sometime".
As groups get larger you end up taking tokens as an IOU aide-memoir. And to nail those who try to cheat.
That then evolves into money. Token money. Tokens representing promises.