“…On the other hand, in view of the previous non-trivial results [20,22,23,24,25,26,27], one can, a contrario, imagine that such a "diagonal reduction" procedure keeps "some" relevant analytical informations, especially on the singularities. Replacing the quite involved multiple integral by an integration on only one angle, and since the integrand is an algebraic function of trigonometric functions of this angle and of the variable w = s/(1 + s 2 )/2, the result will be holonomic.…”