Swing. A rigid rope of length L (pivot to the rider's centre of mass) gives θ'' = -(g/L) sinθ - (ρ CdA L / 2m) |θ'| θ'. This is solved without the small-angle approximation (RK4, 1 ms steps).
Period. For amplitude θ0 the exact period is T = 4 √(L/g) K(sin(θ0/2)) where K is the complete elliptic integral of the first kind (computed by the arithmetic-geometric mean). The familiar 2π√(L/g) is only the limit θ0 → 0; at 120 degrees the real swing is 37% slower.
Speed and tension. Energy gives v2 = 2 g L (cosθ - cosθ0). The rope supplies the centripetal force: T = m (v2/L + g cosθ), so at the bottom T = m g (3 - 2 cosθ0) - 2 weights for a swing from 60 degrees, 3.5 weights from 120. A rope that cannot take it snaps and the rider leaves on the tangent.
Release. After release the rider is a projectile launched tangent to the circle, at height yr above the landing level with speed v(θr). Range x = L sinθr + v cosθr t, with t = [v sinθr + √(v2 sin2θr + 2 g (yr - yland))]/g. Releasing late gives speed but a steep path; releasing early gives a flat path but low speed. The best angle is a compromise, well below 90 degrees, and depends on how far below the release point the ground is - it is not simply 45 degrees.
Gravity cancels. Both v2 and the flight time scale with g, so without drag the range is the same on the Moon as on Earth, only slower. Air drag breaks that symmetry.
Simplifications. Point mass at the CoM; rigid massless rope; no swing-out of the rider's body; drag is CdA v2 (CdA 0.9 m2 default); landing ends the run; the rider is an abstract robot.