Answer:

Depending on the time relationships between waves-a & c and c & d, both methods can be valid. My decision to use one or the other depends on the information they provide.

The most reliable time forecasts are possible when two adjacent, same-degree waves (within the same pattern) take vastly different amounts of time. In that case, the third in the sequence will nearly always be half of the previous two combined. Therefore, if waves-a & c are equal in time, I can only know wave-e will be different in time than a or c, but I won't know if it will be larger or smaller. So, if a & c or similar in time, I'll take a look at waves-c & d to help with a forecast. If c & d are vastly different in time, I'll use that relationship to project the time for wave-e. Occasionally, both relationships can be used to more precisely pinpoint the final time range for wave-e.