I am skeptical of this analysis for several reasons:

1. For better or worse, skilled personnel providing the service would be protected from malpractice to some degree by the very nature of the procedure. It's simply too hard to prove that harm occurred by all possible future standards. Even if malpractice is a possibility, the penalties would tend to be lower.
2. The $25,000 for cryoprotectant sounds unrealistic unless there are no competing firms producing comparable cryoprotectants. That Alcor pays this at present seems to be a result of the fact that M22 is produced and consumed on a very small scale. Nonetheless, an insanely high profit margin for companies developing less toxic cryoprotectants could be a very good thing for cryonics quality. Cryoprotectant toxicity is an incredibly important area to develop for the purpose of not only cryonics but also for organ preservation and the advent of true suspended animation.
3. Responding rapidly to cardiac arrest is something we pay for already with the existing health care system in developed countries that have EMS. If cryonics provides motive to extend EMS infrastructure to undeveloped countries, that is useful for other reasons -- it would save lives in the ordinary sense of the term. It should not be considered an additional expense of cryonics except in situations where EMS would not be deployed.
4. A society that takes cryonics seriously would feel very differently regarding assisted "suicide". It would be something encouraged (and voluntarily pre-arranged) for situations when the brain and personality is seriously threatened. This would result in extraordinary savings not only by eliminating the need for emergency response and standby for cryonics, but coincidentally by reducing the burden (which is intense both financially and psychologically) in terms of caretaking. Not to mention that the patient would actually possibly survive.
5. Monolithic domes are cheap to construct, and well suited for cold storage warehouses. This is not the same thing as cryogenic warehouses; the engineering requirements are more stringent, but their use as a cold storage room at -135C does not seem implausible. The first of such cryogenic warehouses to be constructed would be more expensive than the last due to development requirements for a robust cryogenic storage system. I do agree it would realistically be kept to a smaller scale than national or continental -- municipal seems reasonable. The warehouses would need to have an automated pack-and-place mechanism for easy storage and retrieval, which could be maintained externally. Patients would be stored in crates which can be added in through the top and moved into place by an automated crane. Retrieval of specific patients would require that all crates stored on top or in front of them would need to be moved, but should be feasible if necessary. Other kinds of cryogenic goods could be stored in the same warehouse temporarily in similar crates until enough patients are accumulated for it to pay for itself.
6. Large scale facilities would cost less because they do not need to be kept at liquid nitrogen boiling temperature, only at the glass transition temperature. The heat sink effect of the large amount of thermal mass comprising the structure and other patients would prevent it from fluctuating without the necessity of LN2 boiloff. Patients would also benefit from this in terms of reduced or eliminated cracking effects.
7. The only energy (heat) source in the plant would be the crane. There could however be explosive potential depending on how the building was kept cool. If there is an LN2 pipeline or tank involved, that would be something that needs to be engineered very carefully. This is potentially an argument in favor of (multiple redundant) electric coolers. On the plus side, cryogenic engineering is a fairly mature field and LN2 storage is commonplace at hospitals. I don't think the required level of precautionary spending would be nuclear plant levels.
8. Monolithic domes are extremely stable against tornados, hurricanes, and earthquakes. Additional earthquake proofing could be added below ground to help ensure that patients are not jostled too much, as is done for hospitals, but my guess is we're talking a few million dollars for tens of thousands of patients if not better.
9. You mention halving the neuro storage cost. But logically you are either saying that storage costs would go to half across the board, or that full-body costs go to half that of a neuro. If neuro is 1/7th of full body, this implies that the cost is 1/14th, which is pretty significant.
10. Whether we like it or not, the wealthy, to maintain their place in the status hierarchy, would tend to spend dramatically more than the publicly-funded or middle class, just as is the case with medicine today. Spending several million on a cryopreservation with a tiny chance of working better is not a big deal for a billionaire, any more than buying a private jet would be. To some degree this subsidizes the infrastructure and research for the less privileged classes.

Finally, it bears emphasis that even if you're right and the costs do tend to rise dramatically from the causes you are suggesting, most of them imply much better cryonics. The worst cryonics then would be better than the best cryonics today. Furthermore they would naturally evolve into a suspended animation style cryonics as soon as reversible suspension is developed, ensuring that hardly anyone "dies" of aging, cancer, etc. once that milestone is achieved. As it stands, if true suspended animation were discovered tomorrow, cultural (and financial/organizational) inertia would keep most people who could benefit from signing up for it for decades to come. Thus in terms of cost per unit value delivered, I would still maintain that universal cryonics scales very well.