Simultaneous sampling of disjoint observables

[Sola85]

Required prerequisites

• Search the issue tracker to check if your feature has already been mentioned or rejected in other issues.

Describe the feature

cudaq.observe seems to generate more kernels/programs than necessary. E.g. one of the internal test cases

hamiltonian = (5.907 - 2.1433 * spin.x(0) * spin.x(1) -
                   2.1433 * spin.y(0) * spin.y(1) + 0.21829 * spin.z(0) -
                   6.125 * spin.z(1))

generates 4 kernels that are executed separatly, but 3 should be sufficient, since z(0) and z(1) can be sampled in the same shot.
Is this a feature that I simply have to enable, or is this just not implemented yet?