Improve consistency.detect_energy_generating_cycles

[ChristianLieven]

Problem description

Despite #189 having been closed a long time ago, I had to remove the test for an infinite proton gradient i.e testing if the model can somehow cycle h_c and h_p with closed bounds. This test however is very specfic for the existence of a periplasm and since it is not trivial to presuppose agnostically between which two compartments a given model builds the PMF, I've removed this capability from this test. It should be re-added though once we can identify prokaryotes vs eukaryotes.

It could be possible now that we identify compartments based on an internal mapping. This test could be carried out as long as either the periplasm or the mitochondria can be identified in a model.