Arrogant speculation alert, this is all hypothetical theoretical assumptions.
I would have thought that crystal would behave very differently from metal. Metal tends to harden from repeated stresses and turn brittle. Crystal has a very precise structure and starts off brittle. I would expect that the vast majority of stresses on crystal are a matter of either working or not working, which would make maintenance and operation a very binary matter. You can't really repair crystal without growing it again and machining it again, probably with a fault-line where the old growth misaligns with the new growth. So, basically, repairs are impossible, you need replacements. Conversly, it doesn't really suffer from degenerative failure. It will have its surface grind off if subjected to the correct forces, and chips can be taken out, but I am not familiar with crystals suffering from repeated heating and cooling. Either it expands enough to shatter along a fault, or it doesn't.
I would imagine that we can either make steam-engines of proportionately large volume out of crystal, which don;t really fail without outside help, and lighter due to superior material performance compared to weight, or smaller metal steam engines which can be maintained and repaired but also need repairs and maintenance. So the metal ones would get you more power for the same volume and could be maintained in the field, while the crystal ones would get your more power per weight and last longer but would need to go back to dock when they failed.