Thanks Dave. What will happen will depend upon several variables, among them he amount of water in the boiler, the depth and intensity of the fire, the state of the boiler (heat and pressure reading) before uncovering the crown, the number and size of safety valves, the length of time the crown is uncovered, and then of course the physical condition and design strength of the boiler. Lets say there is a good fire, the kind normally wanted for continuous running hauling passengers, the boiler is at full operating temperature and pressure and water is low but not in the danger zone if in a level position. What is most likely to occur if the crown sheet is uncovered and then quickly recovered is a momentary severe over-pressure which should be released by the safeties. If the crown is left exposed for longer and is allowed to get very hot the safeties probably can't release enough pressure soon enough and the probability is a failure will occur at the weakest point in the area closest to the pressure source.
In the event of such a low water disaster-in-waiting the thing to do is stop the locomotive wherever it is, drop the fire immediately, and walk away (ie, wait) and allow the engine to cool. At that point whatever damage there will be has been done and under no circumstances should an attempt be made to introduce cold water into a hot boiler, even a tiny amount. This will produce an extreme over-pressure failure and a ruined boiler.
Of the books listed above my favorite is the KN Harris book, but what I know about model boilers and boiler building (by far not everyhting there is to know) has been learned from several sources including books, magazines, my mentors, trial and error, and just building them. The photo is of a boiler for a 4.75" gauge (1"scale) tank loco I built for a fellow in Ohio a few years ago.