Nostr has too many features. I’m building a soft fork with no DMs, no zaps, no reactions, no media. Just text. you can zap some random nostr wierdo in my behalf.
And this better isolation + smaller number attempt to retroactively add an asterisk after your "QEC makes the N2 term worse" from earlier, pretty sure you know that can't fly.
Under you assumptions, the error rate for the logical qubit has to *always be worse* than the physical qubit, no matter how good the isolation is. But look, it's actually better. Also the logical lifetime of the logical qubit would have to always be shorter (for Google's Willow it's like 3x longer).
The results prove irrefutably that the N is *not* the governing factor in these QEC systems at all.
To argue that the ceiling remains at 170 despite multiple results showing the logical error rate is better (yes better) and the lifetimes longer (yes longer) than the physical is to break your own math.
It's correctable, exponential scaling, proven out by experiment. Not the uncorrectable, quadratic scaling your math depends on, and that you just broke with your asterisk anyway.
