I followed the link you gave and found the following on page 8-8 of the manual (pretty much (d) confirmed what Morton apparently did...I'll still need to work out the why...). Check this:
(c) It is within this range of torpedo track angles that the greatest amount of course error can be absorbed. From a study of the curves it is evident that the maximum deflection angle is obtained when firing on the optimum torpedo track angle and that the optimum torpedo track angle has a value equal to 90 degrees plus the maximum deflection angle. It is also evident that as the target speed increases for any given torpedo speed the slope of the curves becomes sharper. This means that the higher the target speed the greater the rate of change of deflection angle with torpedo track angle. It is therefore true that the optimum torpedo track angle is more effective for absorbing errors in course when the ratio of torpedo speed to target speed is large. It therefore may be stated that the optimum torpedo track angle is a good mean torpedo track angle for firing a salvo of torpedoes if the target speed is less than one-half of the torpedo speed.
(d) The optimum torpedo track angle for a 16 knot target for a 46 knot torpedo is about 110 degrees and for a 29 knot torpedo about 125 degrees.
So based on this a 120 track seems reasonable. I would have thought something ahead (Cromwell style) would be best, but I suppose it may not be...