It's been a classic dilemma of the Cold War: What should be done early on in an air war?
* Should the air forces focus on the air superiority fight (fighter vs. fighter) at first?
* Should they focus on the destruction of enemy air defences first?
* Maybe attack enemy airfields?
* Maybe blow up some important bridges?
* Attacks on hostile troops on road marches?
* Attacks on hostile troops in contact with friendly troops?
* Should squadrons relocate to more survivable airfields or stay at their home airbases?
* What kind of mix would be optimal?
The Israelis had this kind of dilemma - especially in the surprising Yom Kippur War 1973. They seem to have improvised. The substantial losses of their attack aircraft forced a campaign against hostile battlefield air defences on them in the midst of the conflict. In the end, lots of technological changes and special conditions prevent the Yom Kippur example from giving us reliable guidance about how to answer the dilemma.
The dilemma wasn't nearly as serious in the conflict against Iraq. The Iraq was simply no peer and not capable of immediate decisive action on the ground or air. Fighting against Iraqis was even less than a sparring match in comparison to WW3 expectations. Again, there's little to learn from the campaigns against Iraqi forces in regard to the basic prioritization dilemma.
- - - - -
There might be an answer to the dilemma, though: Surprisingly, this may be a technological answer (and we should be sceptical about it for this reason).
It's obvious that several of the aforementioned options are related to the survivability of combat aircraft. Survivability against hostile fighters, against hostile attack aircraft (when on the ground) and against hostile air defences.
Now what if we were able to take this out of the equation? Let's assume we had a silver bullet that can strike operational level targets (typically 50-500 km depth, for example) while the artillery can strike close targets (and substitute for lacking close air support).
The air forces would then be able to fight for air war superiority, fight air force vs. air force. They would have the best probability of success, could later turn on the hostile ground forces and deliver a strong argument for the politicians who hopefully keep negotiating about an end of the folly.
OK, which weapon or munition could render fighters, air defences and attacks on friendly airfields quite irrelevant? The (quasi-) ballistic missile!
Such missiles are very survivable against most air defence systems, have a useful range for the operational level of war (the longer the range the faster - and thus more survivable!) and nowadays such missiles have the necessary pinpoint accuracy for the destruction of stationary (fixed and reconnoitered semi-mobile) targets: Air fields, long-range air defence batteries, bridges).
NATO air forces (and navies) have understood their potential, their potency as threat - and accordingly spent a great deal of attention and money on hard kill defences against such missiles.
They did not embrace the (quasi-)ballistic missile themselves, though. Missile types with less than 500 km range would fit into the treaties that are in force (except possibly ICOC 3-3).
It may be a prejudice, but maybe it's simply bureaucratic inertia coupled with conservativeness and special interests (fighter pilot generals wanting more fighter wings, not more unsexy missile batteries) that keeps these missiles outside of NATO air forces.
Foreign policy strategy (promotion of ICOC & BM counter-proliferation efforts in general) might play a role as well.
The exposure to Third World ballistic missiles based on Russian 1950's technology has distorted the perception of the (quasi-)ballistic missile threat. Such missiles are at times interpreted as useful only with non-conventional warheads.
It's almost forgotten that NATO had such battlefield missiles with conventional warheads in service during the Cold War!
- - - - -
There are several modern designs of accurate (quasi-)ballistic missiles:
supposedly 300 k range
supposedly 400 km range.
The Israeli LORA missile:
Supposedly 300 km range.
The payload is several hundred kilograms each - enough.
The most obvious choice for NATO forces would probably be to introduce (more) ATACMS Block II into Corps- or Division-level army artillery units and to produce in license a longer-range version of LORA (to be honest, it's most likely cheaper to let them develop a LORA 2 and to buy a license than to develop a missile of our own!).
The dilemma could then be solved quite easily; NATO air forces could alternate between defensive (defence with fighters and air defences) and offensive (additional strike packages against battlefield air defences and relatively easily accessible installations) phases until it has won the air power vs. air power contest in one shape or another.
Strike missions against airfields, fixed and semi-mobile area air defence assets, bridges and the like (ministries?) would be substituted for with the fires from 300-500 km (quasi-)ballistic missile regiments.
Close air support could early on be substituted for with army aviation and artillery fires.
So far, the Western air forces don't seem to believe that this is necessary, though. They prefer air-launched cruise missiles of about 250-300 km range instead. Such cruise missiles require sorties just as the classic strike packages would do.
Maybe we should pay more attention to (quasi-)ballistic missiles as a gap in our air forces instead of paying obedient attention to other big ticket projects (fighters, bombers, air launched cruise missiles) and to the role of ballistic missiles as threats only.
F-35 and Typhoon critics are numerous - how many critical remarks about the lack of SRBMs in Western air forces did you see (except here)?
edit 2013: I should have mentioned the Indian Prithvi missile as well.