Sourpuss,
I have a calculator by my side now, at 210 TAS the radius of turn at 25° bank is 1.4 nm, the 1% of TAS works for 210 Kts at 17° angle of bank.
If it's good enough for PANS-OPS to use a time lag for entry / exit as you've alluded to, then I gues it's OK for we mere mortals. Allowing their 3 seconds Roll In / Roll Out, that adds a further 0.4 of a mile, making the 1.4 nm an actual of 1.8 nm, a mere 0.3 nm short of the perfect asymptotic interception. Actually, this is nice, as it gives a short "normal" intercept time at a shallow angle, or alternatively, provides some protection against an overshoot should you have a Tailwind.
Definately good enough for every-day work, please don't use it if you gain employment writing programmes for FMCs. Better then to use R=V^2/Tan Bank Angle / G.
Regards,
Old Smokey