As indicated above oxidation of an aluminum surface is a self limiting process where the oxidation itself tends to seal off the underlying aluminum from any further oxidation - but only to a point.Basically what Doc Intrepid said. Aluminium is a much more useful metal for scuba tanks because it's lighter, cheaper, and when it does oxidise - the equivalent of rusting - the oxidisation process actually protects the metal underneath. You will notice that brand new scuba tanks are bright and shiny but they oxidise very rapidly and go a dull whitish sort of colour - again, the same process makes aluminium window frames so durable.
If you have water, and in particular salt water, in an aluminum tank it is going to oxidize and fail very quickly (and the same is true for a steel tank).
Aluminum is also lighter than steel but it is not as strong, so the walls of an aluminum tank need to be at least twice as thick - offsetting the weight savings of the lighter metal. Aluminum also has fatigue and life cycle limits that do not apply to steel tanks, and aluminum is much softer and much more prone to damage that may result in it being condemned. So when you look at the bigger picture, aluminum does not look so hot as a tank material despite the closed nature of the oxidation process.
For example:
1. A steel X7-100 is smaller in overall size than an AL 80 and only weighs a pound or so more, but holds an honest 100 cu ft rather than the 77 cu ft in an AL 80.
2. There are 100 year old welding tanks still in service that still pass hydro test and still qualify for plus ratings - not exactly what one would expect from a steel tank that spent 100 years filled with O2. There are as of yet no 100 year old aluminum tanks, but I seriously doubt that there will be many 100 year old AL tanks in active service given the track record of AL tanks to date.
3. In Indian Springs where cached bottles are left in the cave for 2 months at a time, they switched to steel tanks as they held up much better than aluminum tanks when constantly submerged - again not what you would intuitively expect.
Again, a high percentage of O2 by itself does not increase oxidation in an AL tank or rusting in a steel tank. You also need moisture in the tank for the oxidation to occur regardless of the FO2 or PO2.
Now in a steel tank used for O2 service is makes sense to tumble the tank during the O2 cleaning process to remove all traces of rust for three reasons:
1. Rust attracts moisture and will concentrate it against the steel under the rust.
2. Rust is flammable and is a potential fuel source (rusty steel wool makes great kindling)
3. Rust particles, if entrained in the O2 system at high velocity, could cause a park if they impact another surface and consequently provide an ingition source.
Aluminum tanks, on average, do not require quite as much prep for O2 service.
As for cost, hot rolled steel plate currently sells for about $.40 to $.58 per pound while raw aluminum costs about $.65 per pound, so even when you correct for the inevitably higher costs of properly alloying both metals, the cost of steel is still cheaper per pound than aluminum and the number of pounds per tank is very close so the material cost difference is minimal.
-----
Contrary to popular opinion breathing a tank down to the point where you cannnot get any more will not result in water entering the tank. Very few if any regs will breathe a tank down to ambient pressure and even if one did, water could not flow upstream unless the lever in the second stage were depressed. So in the real world you would have to breathe the tank down as far as possible under water, then depress the purge to open the system and then go deeper by another 50 - 100 feet to increase the ambient pressure enough to overcome the pressure still remaining in the tank. That is quite frankly just not going to happen in the real world. I'm surprised Myth Busters has not busted that one yet.
What will result in water getting into a tank is a compressor with a faulty moisture separator and/or filter that does not properly dry the air. The other way water commonly gets into a scuba tank is by water being present in the valve or in the fill whip when the tank is filled. If the fill operator does not crack the valves on both the tank and fill whip to blow out any moisture that may be in them, that moistrue will be blwon in the tank during the fill. Both are dive shop related failures so it is somewhat ironic that dive shop associated instructors still preach the "don't breathe the tank donw to zero" line.