Why would you not value fiat gains or gains of any kind?

The sampling of yield consideration isnt intended for strictly fiat gains. That angle is only possible in the present due to lender of last resort.

The goal is increasing ones wealth overall, as measured in Bitcoin terms. If you start with 100000 sats, and earn either 5000 sats directly (5%) or earn an equivalent of 5000 sats in fiat terms beyond the delta in fiat value in the sats lent then you still earn, presuming you get a return of the interest and the principal.  Obviously any lending and counterparties represents a risk, thats what interest on that lending is for and why seeking such isnt for everyone. 

Saylor's hypothetical is based on assuming we still have the dollar, we still have the USG and the Fed and the big banks. He's working within the system to speculatively attack the dollar ala Pierre Rochard (2014), but as time goes on he looks for more ways to grow, or cover expenses without selling. Hence, some yield, on a portion may be acceptable depending on risk.

Ammous perspective lies more in Bitcoin Standard extremism. No USG or lender of last resort. No JP Morgan. Somehow loans are either bad, or at virtually 0% interest (because it costs more to hold an asset then the risk lending it out). I just felt like Saif was reaching hard in this one for some mental gymnastics that make huge assumptions on the future from weak data points. He brought up the book A History of Interest Rates, and a chart showing long term downtrend but neither him nor that book adequately cover reasons for declining rates.

I didnt participate in the BlockFi/Celsius/FTX fiasco. I lean risk averse. But Im not opposed to leveraging portions of wealth in an effort to garner more return or pull value forward. This is why people buy houses with mortgages instead of paying in cash. On a Bitcoin Standard, a loan would still be viable provided the earning or funds flow is verifiable.