A commenter on my last post on this topic pointed me to a search that uncovered two interesting and helpful papers I was not aware of.
This review traces the history of methods distinguishing between VT and aberrancy on the surface electrocardiogram. The tried and true method, which assesses for AV dissociation (including that evidenced by fusion and non aberrant capture) or VA conduction with block, is accurate and conceptually sound. However, it is difficult to use with faster ventricular rates, at which atrial impulses are often not well seen. The use of morphologic analysis has gradually improved. Unreliability of early criteria drove the emphasis on analysis of AV relationships. Simple criteria based on QRS width or axis have fallen out of favor. Various early criteria were based on the resemblance of the QRS complex to either bundle branch block. These were popular because they were intuitive, but suffered from low sensitivity and specificity. Several improved methods were developed leading up to the popular Brugada criteria which established morphologic analysis as a more reliable method. More recently criteria involving only lead aVR have been validated.
The aVR method, as explained in this paper, is comparable in accuracy to the Brugada method and appealing in its simplicity.