There was very nearly a fatal at Scampton when a USAF exchange office decided to attempt a 2-engine simulated asymmetric overshoot from below VCH...... They went sideways across the aerodrome until the other engines spooled up.
It was explained to us at our UAS (back in the days when real RAF QFIs imparted wisdom to youngsters during UAS training nights, rather than relying upon the "Janet and John Go Flying" books of later years), that the reason for many crashes in the Meatbox/Vampire era was due to acceleration errors affecting their steam-driven artificial horizons. These gave a false pitch up and roll right indication, hence Meteor-shaped holes appeared off the left hand side of many runways at the departure end. To compensate, 'pendulous vanes' were fitted to the air-driven gyro suspension, these opened under acceleration and applied corrective air jet forces to eliminate the 'Meteor hole' effect. "About the only time pilots were glad to have the company of pendulous veins (or rather, vanes)" quipped the CFI!
Asymmetric practice probably killed more people than actual asymmetric landings; I did 3 trips in the back of Meteor T7 'Clementine' at Brawdy in the mid-70s (being shot at by fellow Hunter students), and on each occasion a very positive asymmetric brief was given. For one pilot, with rather little legs, this included assisting him with full rudder as called. But back in the 1950s, things weren't quite so cautious, by all accounts.