Maintaining flight in a 90 degree bank is something aerobatics aircraft can do, but it's hard, requires a specific aircraft setup, and you're not doing it at approach speed. More like 100-150mph.
Very easy spin an aircraft when attempting to maneuver at low speeds, and that will invariably result in a crash as you need several thousand feet of altitude to recover. Such crashes are 99.9% fatal.
Usually looks something like this (Warning: It's a crash, there's a fireball, but it isn't overly graphic): https://youtube.com/shorts/urTs-y7MiJE?si=ovfiC3ZvCgcFr5Nu
Rather than just rolling out the pilot pulled back sharply right around when when he got fully inverted, you see the nose quickly start to come around, and then a split second later the left wing stalls deeper at which point it enters a counterclockwise spin and impacts the ground after approximately a quarter rotation.
| During official trials, the Tabloid achieved a maximum speed of 92 mph, with a stalling speed of 36.9 mph.
https://www.ctie.monash.edu/hargrave/sopwith3.html
You have a lot of good points above but missed out of a great way to reduce stall speed: less weight! As you burn off fuel or otherwise shed weight, your wing loading decreases and thus you are able to decrease your angle of attack for a particular airspeed. Of course, angle of attack is the key to a stall (and thus a spin) so we can talk about other ways of reducing the angle of attack, often by unloading the wing.
Eg: another way to unload the wing is to reduce the tail-downforce that the main wing needs to counter in this classic design. By moving the center of gravity as far aft as is practicable, you achieve the same goal. As with anything in aviation, this is a trade off; By moving the CG aft, you make recovering from a stall harder or even impossible. I imagine that stalling in one of these bi-planes couldn’t have been a benign experience to begin with so, pilot beware!
Weight doesn’t help nearly as much as you’d think. Lift needed scales linearly with weight but lift scales exponentially with speed
ie: e^ax behaves very different than x^a
Lift is proportional to a*v^2 where “a” is a bunch of constants based on wing geometry, density altitude, etc. v^2 is most definitely not exponential behavior.
I'm not sure that's the quote. You can't after all eat your own head, even once.