OAG,
My mistake; I realised after I posted that bank angle becomes defined using the technique but did not have time to edit it. One point to consider using the "zero wheel" techniqus is that it works well for jet aircraft with swept wings whereby most of the rolling moment under asymmetric power is as a result of sideslip and static lateral stability. However, in twin (or 4-engined!) propeller aircraft, an engine failure results in a significant loss of lift over the wing that has the failed engine, due to the loss of propwash, thus giving a strong rolling moment. Combine this with the fact that static lateral stability will probably be less than on a jet airliner (straight wing vs. swept wing) and the use of rudder to achieve zero wheel in a prop aircraft may result in a large sideslip angle as you are using the rudder to generate sideslip and thus create a rolling moment to oppose the lateral lift asymmetry. It would be interesting to get the views of some twin turbo-prop operators on the "zero wheel" technique.
The 5 deg bank angle question is an interesting one, and the precise AoB to achieve stabilised flight for a given sideslip angle is mainly a function of the sideforce characteristics of the aircraft (if not at zero sideslip) and the lateral aerodynamic force produced by the rudder. I suspect that 5 deg is generally considered as most aircraft have bank angle markers on the AI at 0 and 10 deg and 5 is then relatively easy to judge; 2-3 deg would not be so easy to judge. Using too much bank (10 deg) would give a significant reduction in the vertical component of lift, necessitating an increase in AoA and hence drag to maintain stabilised flight. The extra drag would then degrade climb performance so this would be inefficient and counter-productive.
One final thought on this thread is that the theory of flight under asymmetric power, with respect to rudder and lateral control inputs, leads to an optimum solution that will give the best controllability and performance for virtually all aircraft. However, this theoretical solution may be difficult to fly, and for any specific type there may be a procedure that is easier to teach and to fly that still gives adequate performance and controllability. Hence the differences in techniques discussed by the various respondents in this thread. Remember, fly what works in your aircraft but accept that this may not work for other types and you may need to learn and practise different techniques.