Yield doesn’t exclusively come from the fact there is counterparty risk though. There is cost to deferring consumption, and providing the security and matchmaking so whilst banks needed vaults and guards for security (you could do this yourself - private credit has always been a thing), Bitcoin relies on channels which also need security and the matchmaking is routing. That’s why the Austrian definition is broader than purely having counterparty risk and considers duration and other costs for why interest is not inherently bad.

Price appreciation isn’t yield because 1 BTC = 1 BTC, but 1 BTC in a Lightning channel should over time = >1 BTC with a risk that it = <1 BTC. I agree it’s fees in the purest sense rather than “yield” but I feel like this is semantics and not really productive. 

Bitcoin doesn’t map one-to-one with legacy concepts so when we apply them they don’t always fit neatly and that’s a big part of where people go wrong. It’s why Saylor is getting it wrong. We have to work within the rules of Bitcoin, imposing outside ideas on it only helps us in the abstract but if our ideas don’t exist within those rules then they’re rough proxies at best and this is an example of that.