Common Fallacies in Cosmic Logic

The following type of argument is commonly made by cosmologists:
"Observation X falsifies MOND, therefore dark matter exists."
See, for example, commentary by such luminaries as Ned Wright and Sean Carroll.
As of this writing, the most popular observation X = the Bullet cluster. I concur - the mass content of the Bullet cluster is more naturally explained by dark matter. However, the opposite situation applies to the remarkably high collision velocity of the self-same Bullet cluster: that is more naturally explained by MOND. I hear that mentioned a lot less.

It is a logical fallacy to presume that the failure of one theory guarantees the success of another.

Our familiar but frankenstein-ugly cosmology, ΛCDM, requires the existence of two novel entities: dark energy (whatever that is) and dark matter. The latter can't be just any form of unseen mass. It has to be non-baryonic, dynamically cold dark matter. The widespread presumption is that this is some kind of new particle, specifically a WIMP (Weakly Interacting Massive Particle) from the super-symmetric sector (something like a neutralino). We should bear in mind that supersymmetry remains an attractive but unsubstantiated hypothesis for what might happen beyond the resolutely successful Standard Model of particle physics. Until we actually detect these hypothetical particles in the laboratory, I don't think we can claim to know they actually exist. If they don't, ΛCDM collapses... a fate that no more guarantees MOND is right than problems with MOND guarantee ΛCDM is correct.

Nevertheless, we have become so familiar with non-baryonic CDM as a dark matter candidate that it is easy to make the following leap of logic: The Bullet cluster requires unseen mass even in MOND. Therefore CDM must exist. Well, not quite. Unseen mass is required in clusters in MOND, a major detraction for a theory that would explain cosmic dynamics without appeal to invisible mass. However, we don't know what the unseen mass is. Unseen mass is not guaranteed to be CDM. Some have suggested it could be heavy neutrinos, a hypothesis I am not comfortable with. There are plenty of missing baryons in the universe, presuming big bang nucleosynthesis is correct (which is entirely consistent with MOND). Those could easily make up the residual MONDian mass discrepancy in clusters. I'm not comfortable with this hypothesis either because I don't know how to hide so many baryons. But strictly speaking, MOND has a missing baryon problem in clusters. So does ΛCDM. At no scale, including clusters, does the detected number of baryons add up to the amount expected from the cosmic baryon fraction. If both theories suffer a missing baryon problem, well, they both suffer a problem.

I guess what I find most unfortunate is the following fallacy:
MOND is wrong because [insert your reason here], therefore I don't need to think about it.
Let us accept for the moment that MOND is indeed falsified. Even in this case, it has been quite well established that MOND provides an economical mathematical description of rotation curve data. That must be telling us something. At the very least it is an organizing principle that dark matter obeys, if only in rotating disk galaxies. Consequently,
The observed MONDian phenomenology requires an explanation in the context of ΛCDM.
Unfortunately, I see no discussion of this critical point. I have spent an inordinate amount of research time trying to address this very question. I have not published a single thing on it because I have never come up with anything remotely satisfactory. This is a hard problem, but an essential one to crack if ΛCDM is really the correct picture.