Ignoring the energy component of the crystal mining station's life-cycle cost may be fine if you want to keep your calculation simple, but if you are trying to follow a process of risk analysis and return maximisation, your resulting calculation will be less precise and less accurate. You're right that energy maintenance applies to every other mining station, however assuming you are able to protect them militarily, the risk of not getting the return on your investment is negligible.
The question really is 'should you build on 1, 2, 3, 4, or all' of the crystal asteroids. I think the answer is probably 3 for maximal probable return but it might be 2 - the other question is 'do you want to exploit 3 faster and riskier or 2 slower but safer'
Even if your fleet is ready to clean out the hive it may not make sense to build the fourth - unless you park the fleet in the system while you build them all and wait.