My above method of course estimation is not all that good as I hoped I'm afraid
It worked well a couple of times because the plotted bearing lines where almost in parallel to each other. But if they are divergating then my suggestion of picking an arbitrary point on the first line is not correct.
At least this method can be refined a bit in this way:
First draw the first bearing line and track your movement as well as letting the stopwatch run until you draw the second bearing line. And at this point you should be able to get a range estimation via stadimeter for example. Then you have one fixed point on the second bearing line and with that and the estimated length of the targets line of travel you can eventually draw this one and read the target's course off of it.