While I am no expert on aerospace assemblies and coatings. I would expect that these carbon dioxide bubbles would react to the changes in atmospheric pressures and eventually rupture, this would cause the PCB to be exposed where ever these fractures occurred, causing to potential for board and assembly failures.

I don't pretend to be an expert on this; however there are several reasons for cause for alarm in my view i.e. Carbonic acid forms in the presence of Carbon Dioxide and moisture.

Carbonic acid is a weak acid that could cause the potential for failure of the PCB. Further conformal coatings are usually applied to act as a barrier to prevent moisture from the PCB, whereas it seems that moisture may already be present and encapsulated in the PCB, reacting with the conformal coating, and causing delaminating of the coating.

I would insure that the materials are compatible and that the boards are baked or processed to remove moisture before the coating is applied.

Not 100% certain of the details on this particular urethane material but we have seen issues when conformal coating is cured using Convection Heating. In particular, convection can "skin over" the surface of the coating in the early stages of the curing process. When the bubbles try to escape they either get stuck under the skin or break through and create small crater-like pock marks on the surface.

However, when using an Infra-Red heat source, the heating is basically the opposite. The material is heated from the inside out and the bubbles are allowed to escape first and the outer surface is skinned over toward the end of the process. The result is a super smooth surface.

You may want to try a sample or two in an IR based oven to see if this has an impact.

Oh, and to answer your question, we cannot comment on the long term effects of the trapped bubbles but maybe they can be made to go away.