The ephedra hypothesis for soma and haoma has the strongest historical thread and the weakest experiential fit, and holding both facts together is the honest position.
In its favour: the ritual is continuous. Zoroastrian priests have performed the Yasna with ephedra without a recorded break, and a tradition that never lapsed is a better guide to what was originally used than a reconstruction from poetry. The Central Asian archaeology adds independent support. And the linguistic connection between Avestan haoma and Vedic soma is secure, so whatever one was, the other was closely related.
Against it: ephedra is a stimulant. It produces alertness, a raised heart rate and appetite suppression. The Vedic hymns to soma describe expansion, illumination, immortality and a god entering the body, and while poets exaggerate, the gap between that and a decongestant is large. Critics argue that ephedra is a substitute, adopted after the original plant became unavailable when the Indo-Iranians moved out of its range, and that substitution is well documented in ritual practice generally.
The counter-argument is that the Vedic descriptions may be liturgical rather than pharmacological, praising the ritual and the god rather than reporting a drug effect, in which case a mild plant is entirely sufficient and the whole search for a potent one is misconceived.
That last possibility is the one most often left out, and it may be the strongest.