I made reference earlier to 'Hume’s Missing Shade of Blue Re-viewed' by John O. Nelson in Hume Studies Volume XV Number 2 (November 1989).
I think I agree with the following paragraph, which it seems to me is a fair basis on which to consider the case of the missing shade of blue as being not fatal to Hume's philosophy as set out in the Treatise. Indeed, when seen in the context described below, the missing shade of blue is not perhaps the 'exception that proves the rule' but rather 'the exception that proves the nature of the rule.
"One might profitably compare Hume's views in the present regard with Descartes' on essence and existence. According to Descartes, essence and existence are always, at least insofar as God is not concerned,“separable.” Thus, in imagining a triangle (says Descartes) he conceives its essence; nonetheless, ”there may nowhere in the world be such a figure outside my thought." In spite of some appearance to the contrary, Hume, I want to say, similarly maintains that ideas and their relations are separable from existence. Thus, what we discover in a thought experiment, like the color concession under discussion, really says nothing about what exists or does not exist in reality; it at most proves that a certain proposition, namely that stating that all simple ideas have correspondent, precedent impressions, is not an apriori, necessary proposition but a synthetic, aposteriori proposition. On the other hand, the proposition stating that all simple ideas have correspondent, precedent impressions, or the first principle of the science of man, does refer to what exists or does not exist in actual reality. Hence, being only in idea or in one’s thought experiment a phenomenon, the phenomenon of the missing shade of blue projects merely the idea of a contradiction to the first principle of the science of man, not an instance of an actual exception to the first principle and, hence, not something actually falsifying that principle. Thus, without real inconsistency but only an apparent one, Hume is able both to maintain that all simple ideas have correspondent, precedent impressions and that a contradictory phenomenon can be conceived."