I think my analysis was wrong as were the results. An explain is below:
It is an affair of tangent, not cosine.
So the L/D ratio is 2.4 which is Tan(elevation angle of 67°).
Here the difference is tiny, but some other differences are larger as shown below.
As lift to drag ratio is 2.4, drag = 100 N.
240/166 = about 1.445, so Tan(55°30’). The elevation angle is 55 degrees 30 minutes, so far higher. Now, by taking account of the lightly corrected L/D ratio of 4, we have 240/170 = about 1.41, so Tan(54°40’). The elevation angle is now 54 degrees 40 minutes.
384/223.6 = about 1.717, so about Tan(59°40’).
1.22 = Tan(50°40’). Here the difference is very large.
For that L/D ratio should be 0.7, so Tan(35°), not 1.22.
If we keep the initial (lightly corrected) numbers for the kite (240 N lift, 100 N drag, 4 m², lift coefficient of 1, so a drag coefficient of 0.416), and the initial numbers for the turbine drag (70 N, 1 m², drag coefficient of 1), we obtain a L/D ratio of 0.7 by the following way:
0.7 = about 240/340; we withdraw 100 N of kite drag; 240 N of drag turbine remain, so the drag of more than 3 turbines (3.428).
Of course with lift coefficient of 1.6, and yet more with that of 2.9, and drag coefficient x 1.6 then x 2.9 respectively, still more turbines could be lifted with the kite of same area (4 m²), at an elevation angle of 35 degrees:
with lift coefficient of 1.6: 384/548, so more than 5 turbines (5.54);
with lift coefficient of 2.9: 696/994, so 10 turbines.
These results are far from the previous.