Getting away from "platformism" for a moment, I amm thinking the real reson that the Army believes the air Force can't be trusted to do CAS has more to do with availability, timeliness and accountability.
When *we* want to engage a target with our own weapons, it is quite easy; issue a fire control order and engage the target with our own weapons. The fire comes moments after the order, and if there is no or ineffective fire we will know why right away (and often be able to correct it). Even Combat Teams and formations can bring fire "on call" quite quickly through artillery, since *we* own it. Not so much for CAS, since the aircraft might not be in the area, might not have the proper munitions loaded for the support we want and might be tasked by outside agencies on missions we know nothing about. Since things can go south quite quickly if the requested fire does not arrive, it leave a bad taste in the mouth.
IOT regain the availability, timeliness and accountability, the Army needs to re-invest in its own integral fire support. ATGMs like Javelin and Spike provide a means at platoon and company level (especially Spike since it is a FOG-M and can have a man in the loop to identify difficult targets and make the strike/no strike decisions if needed). Mortars with skilled crews and MFC's can provide not only a weight of fire, but also have a much higher ability to hit difficult targets than ever before (especially if technologies like mortar rounds with seeker heads become more available). And the number of actual guns in the artillery park need to grow as well, to provide overlapping coverage and service more targets more rapidly. If UCAV's can be small, smart and cheap enough, using them to identify targets and perhaps deal with them on their own is also a future option.
Some classes of weapons even overlap to an extent. Spike covers an entire family of FOG-M weapons (mini Spike has a range of 1300m, while at the other end Spike-ER can reach out to 8000m, and Spike NLOS is an artillery weapon that has a range of 25km). Since they are ATGM's, they can deal with hardened targets like bunkers, fortified buildings and other difficult targets in addition to tanks. Their "man in the loop" capabilities let you use them "danger close" and as a bonus they also have some ability to discourage armed helicopters. Our procuremenrt system needs to be overhauled as well. Modern electronics are cheap (the real cost of an iPhone or Android device is in the $4-600 range retail), so there is quite a bit of scope for thinking smart weapons can come down a lot in price (imagine if Excalibur rounds cost $7000 instead of $70,000 apiece). This might mean long production runs to gain economies of scale, but having lots of these weapons also means more rounds for training, increased proficiency etc.
As for CAS, if *we* can cover off more tasks with our own integral firepower, then aircraft can be assigned to higher priority missions which play to their strengths (mostly being able to strike deep targets far beyond the range of our weapons, and the flexibility to add a *lot* of weight of fire when the commander needs it.)
