Chris,
In life, I deal with industrial control system's integration. Lots of control voltages (24VDC) for devices, interlocks, and logic; some 3ph power switching and monitoring - - but way shy on electrical theory. Along the way I've had my head in more electrical control panels than I care to count, and (I hope) picked up a basic understanding of electrical applications (especially safety). So all in - - I've never measured current in the neutral of a circuit like Patrick describes. Can't even swear how the hot legs in typical 240V USA-household power are 'phased' - it's never been important. Neither do I know for certain the result of 'combining' 2 phases in a single wire (...saying it like that makes it sound remarkably like a short
).
You've certainly made me think about it and I will have to fall back on the basic application of power here. In the worst case scenario I posed, assuming PF=1, 120V x 20A x 1 = 2.4kW, so the hot wires HAVE to be carrying a combined 4.8kW of power. The neutral has to carry this as well.
In spite of my ignorance, I am betting Patrick's switch-scheme is a no-no in any code in use today. ...I wouldn't do it in any situation I am involved in - personal or professional.