At Astronautix, it says that:

The propellant combinations WFNA/ JP-4 and later IRFNA/JP-4 were the first storable systems given serious consideration in the United States. Problems which caused the abandoning of these propellants were the absence of reliable hypergolic ignition and unstable combustion. IRFNA/UDMH and IRFNA/JP-X finally did prove satisfactory.

By the late 1950's it was apparent that N2O4 by itself was a better oxidiser. Therefore nitric acid was almost entirely replaced by pure N2O4 in storable liquid fuel rocket engines developed after 1960.

Apparently it was so apparent that they have no need to explain why it was so apparent. What was the benefit of going from nitric acid to tetroxide for hypergolics? What was "better" about it? Anyone know?

N2O4 is very stable and can sit in tanks for a long time without ill effect. Pure HNO3 attacks just about any material (stainless steel is safe, but few others are). You can "inhibit" HNO3 by adding stuff (I don't know what exactly) to it, but it makes some nasty compounds when it burns, and I suspect the Air Force and NASA are leery of the EPA.

You get a (slightly) higher Isp from N204. To prove that requires looking up the heat of the reactions and I don't have those materials here. This is partially balanced by HN03 having lower density.

Inhibition involves adding a fraction of a percent of hydrofluoric acid, to form protective fluoride coatings on exposed surfaces.

From Sutton 7th edition (Rocket Propulsion elements, ch 7.2) "There are several types of nitric acid mixtures that have been used as oxidizers between 1940 and 1965...The most common type, Red Fuming Nitric Acid (RFNA) consists of concentrated HNO3 that contains between 5-20% dissolved nitrogen dioxide (N02). The evaporating red fumes are exceedingly annoying and poisonous. Compared to concentrated nitric acid (called white fuming nitric acid), RFNA is more energetic, more stable in storage, and less corrosive to many tank materials. Nitric acid is highly corrosive. Only certain types of stainless steel, gold, and a few other materials are satisfactory as storage containers or pipeline materials. A small addition of flourine ion (less than 1% of HF) inhibits the nitric acid, causes a flourine layer to form on the wall, and reduces corrosion in many metals. It is called inhibited RFNA (IRFNA). Lime and alkali metals are common neutralizing agents in case of spilling, however, nitrates formed by neutralization are also oxidizing agents themselves and must be handled accordingly."

So, basically, it sux to handle and gives only a very slight advantage in density for lower stages over N2O4.

I think N2O4 is actually a little less dense than HNO3, but it could be that my data was taken at different temperatures. Anyway the extra O tells the story. I'm pretty sure that if we work out the LHV (lower heating value) as suggested above we would find that the extra O would make the difference since the energy involved in the CO2 and H2O reactions overwhelm other reactions.

John D. Clark, in "Ignition", writes:

The situation today [1972], then, is this: For tactical missiles, where the freezing point of the propellants matters, IRFNA type III-A is the oxidizer....

For strategic missiles, which are fired from hardened--and heated--sites, N2O4, with a somewhat greater performance, is the oxidizer used.

He also details the history of the addition of 0.6% HF to form the protective fluoride coating inside the tank to protect it from RFNA.

Also, there was a problem with RFNA and Titanium in late 1953 when a technician at Edwards was killed by explosion/asphyxiation. (Intergranular corrosion produced a fine black powder of metallic Ti, which when wet with RFNA was as sensitive as nitroglycerine or mercury fulminate.) N2O4, when anhydrous, doesn't have corrosion problems.

Clark, John D., Ignition! An Informal History of Liquid Rocket Propellants, 1972, Rutgers University Press 662.666 CLAR, Chapter 4 pgs. 47-65