It is important to distinguish between "dwell angle" and "dwell time." Dwell angle remains fixed after adjustment, within the bounds of normal variation that some distributors have (non-electronic ones), but dwell time changes as engine speed changes. Dwell time is the time period that elapses while the distributor cam rotates through the dwell angle. Dwell time is actually the important factor in ignition system performance, but dwell angle is easier to measure, so that is the published figure. The connection between dwell time and the ignition system is the coil saturation time, which is the time period required for the ignition coil primary winding to develop a magnetic field that is suitably strong to fire the spark plugs. That time period is fixed by the coil primary winding (primary inductance), but varies with coil design. Racing coils have fewer primary windings (lower primary inductance) and therefore saturate faster than those with higher primary inductance (a "standard" coil). So a larger dwell angle actually provides a longer dwell time, which permits a stronger magnetic field to develop in the coil primary winding at high engine speeds, the result being satisfactory ignition system performance at high speed.

Enter the dual point idea: since point arcing/gap and coil saturation tend to be opposing factors (a large gap means short dwell angle/dwell time), a second set of points in parallel with the first accomplishes the increased dwell angle without sacrificing the point gap necesary to prevent flash-over with the primary magnetic field collapse. Dual points are usually offset from one another by some number of degrees, which results in the increased dwell angle (dwell time) necessary for high speed operation- that is to say above about 6000 RPM engine speed. A typical dual point, 8-cylinder distributor provides about 40° of useable dwell angle, up from a typical maximum of about 35°. That doesn't sound like much gain, but combine that with a low-inductance (low resistance) primary coil and the current-carrying capacity of parallel points, and you have a good high-speed system. I have always used a distributor machine when setting dual points, but the method described by Joker will work also.

On the topic of timing with a vacuum gauge, I have found it best to use a tachmeter with the vacuum gauge. The process I use is to advance the timing until one or the other reaches a peak, then retard the timing for a loss of 1"Hg vacuum. I have found that leaving the timing at the peak vacuum position leaves the timing substantially advanced. I have found the same result by using only RPM as the indicator. I have also found that RPM and vacuum rarely reach a peak at the same point of timing advance, so when one of them peaks, it doesn't matter which, I retard the timing for the 1" vacuum drop.

I guess this is all for academic interest, for whatever its worth.

Harvester