basically you're removing hydrocarbons (oil!) from the tank walls and, more importantly, the valve surfaces.
The truth is that you can transfill pure O2 into just about anything IF you do it slowly enough. What causes the "boom" to occur is localized heating, which then causes the ignition point to be reached.
When you actually do the O2 add, assuming you're using a proper whip with a needle valve, the O2 going into the tank is REAL cold. Adiabatic expansion insures that is the case; the high pressure gas expands as it leaves the needle valve, which cools it dramatically. The needle valve gets VERY cold - the gas downstream of it is too!
Don't believe me? Do a transfill with an accurate DIGITAL pressure gauge, shut the needle valve, and watch. Heh heh - the pressure comes UP as the gas comes up to room temperature! How's that? Its 'cause the O2 you put in there is COLD and as it warms up the gas expands.....
The trick is to (1) fill VERY slowly, and (2) insure that there are no small restrictions which can provide a heating point - such as the surfaces of the tank valve innards.
With that said, cleaning the tank is not a bad idea. But more important than the tank is the valve itself; I'd tear that down rather than clean the tank, because if you're going to get a "boom" its nearly certain that the valve will be the cause.
BTW, the related chickdiver incident probably was due to adiabatic heating from slamming open the valve. A bad practice, even if we get away with it most of the time. Even a properly-cleaned valve will combust if you get unlucky doing that, as the seat material can, under the right conditions, ignite. O2 tanks have warnings all over them about opening valves SLOWLY, and that's why.
There is no way to know if the valve was actually contaminated from the factory (she related that it was not actually cleaned) or whether she just got unlucky cranking it open to check it. Note that some valves are more dangerous than others in this regard; valves that are "fast-acting" are the worst, while those that are "slow-acting" are the best, as it is the delta in pressure over time that causes the problem.
(Note that for all intents and purposes any FO2 of 50% or higher may as well be pure O2. NASA has tested this pretty exensively and they found that at FO2s of 50% you may as well treat the gas as pure O2 for handling purposes. This means that your MOD 70 deco bottle of 50/50 has the same issues in terms of handling and cleaning as does the 20' bottle of 100%.)
Virtually all currently-produced valves are made with O2-compatable grease (Krytox or Christolube), and odds are that they're perfectly fine from the factory. In fact, I have yet to tear down a new scuba valve that had any sign of contamination under a black light or was lubed with silicone - and I've done a few, as I blend my own gas and as such its MY tail if I'm wrong about this.
Checking and possibly cleaning tanks periodically is a good idea, but the issue is unlikely to be hydrocarbon contamination and a "boom" or even a "fizz" in the tank itself. Its more likely to be progressive damage to the tank walls, especially steel tanks, or valve contamination - and of course you have to remove the valve to clean it, so you may as well check the tank out at the same time.
