[BillCar]

Fixed line is easier than it is made out to be – as joegrundman correctly said, the slower one moves, the more accurate one's reading, but within firing range, I have taken fixed-line readings at flank speed when necessary, and one can easily get away with just doubling the value.

To cop out to the argument of what happened in reality, U-boat commanders would generally use the fixed-line method while having only rough values for the ship's length. These cause a broader deviation in the results than does the doubling of the ship's value, but even when taken together, these two inaccurate estimates aren't particularly bad, and generally proved to be quite adequate when within torpedo range.

To use an example of a tramp steamer crossing your line in 24 seconds:

1.94 x 78m / 24s = 6.305 knots
2 x 78m / 24s = 6.5 knots
1.94 x 75 (as a common estimated length) / 24s = 6.06 knots
2 x 75 (as a common estimated length) / 24s = 6.25 knots

In the example above, all of these values fall within .299 knots of the actual speed. Simply doubling the value by 2 as opposed to 1.94 falls within .195 knots of the actual speed, on the outside.

In my opinion, if one is far enough away for one's firing solution to fall apart based on these sorts of small deviations, one needs to be closer anyway.