That's not really correct. Austrians hold that economic value is subjective. The regression theorem is simply an inference about the emergence of money from barter. It's saying that a medium of exchange must have been valued for itself by the market prior to being adopted for use as a money. From the point of view of monetary theorists, it was necessary to break the perceived infinite regression of saying that something always had value as a medium of exchange. Obviously you have to close the loop somewhere, so this is just a logical inference which keeps you out of the circular logic of saying that something always had medium-of-exchange value when discussing the emergence of money.

Peter unfortunately does use the term "intrinsic value," but he's not an academic. What he means is market value apart from its use as a medium of exchange. What people in the hard money crowd find it hard to grasp is that human beings can quite easily ascribe value to something that was initially valueless (no market price) like the tokens produced by the Bitcoin software on Day 1 of its launch. No bickering about monetary theory can contradict what we actually observe in the marketplace. We see clearly that people started valuing Bitcoin for itself due to its perceived utility or novelty or whatever. It's irrelevant what the reason to value it was in the minds of the participants. The relevant point is that they then form the market for Bitcoin, and it's off to the races. Bitcoin is being used as a medium of exchange now, and so it obviously had to have had a value to people (if not a price) before it started being traded for pizzas and such. That's the point of the theorem. It's expressing the idea that this is a logical necessity. But the regression theorem should not be used as a predictor. It just says that if something becomes a medium of exchange then it was valued for itself immediately prior to that. People tried to use it as a predictor when it came to Bitcoin and got confused.

So one persistent error is in thinking that if something did not always have a market price (like the original BTC tokens), then it was never really valued for itself in the marketplace, and therefore can't become a money (due to the regression theorem). This is obviously a misunderstanding of the theorem, but it's hard to catch when you've focused on the virtues of hard money for so long. We never think of gold and silver as being valueless at one time because most pat explanations of the emergence of money start with precious metals being valued commodities already. But originally they did not have prices at all - just like Bitcoin. They were newly discovered curiosities at one point too. So this failure to go back to the beginning of the story of precious metals led to confusion when people analyzed Bitcoin as being merely valueless tokens with "no intrinsic value." Add to that the fact that Bitcoin was engineered on a computer and you get an understandable resistance in people who have been immersed in classical monetary theory.