Okay, soyou do need a dedicated CAS platform, who says it has to be an A-10 though?
That story is apocryphal, they were among the first aircraft to deploy to Saudi Arabia after the Iraqis rolled into Kuwait. Plus, his son flew Warthogs in Desert Storm and it would have made for an awkward Thanksgiving if he had screwed his boy out of a chance to go to war.
And sometimes you just need a random stick, not even a jet. An A-29 Super Tucano can do many of the things an A-10 does in a low-intensity conflict for a fraction of the cost so why not use those instead? Do you really need a 30mm anti-tank cannon and a crapload of armor to take on the Taliban or ISIS?
Crapload of armor? More like a metric butt-ton of armor. Along with dual hydraulic flight controls with manual reversion and assorted other features that will allow it maul the enemy, take some punches, survive and return the pilot in a LIC. The political ramifications of a captured pilot are very problematic when dealing with non-state actors.
As for the cannon, to paraphrase from the Army, heavy can go light when required, but seldom can light go heavy. This really gets down to philosophy. I personally believe the single worst thing of all in on the F-35 was the premature shutdown of the F-22 - a no compromise Air Dominance machine. There is a need for a no compromise, spare no expense, latest tech fighter - a Ferrari. There is also a need for a low tech, low expense, heavily armored F-350 duallie. I tend to think that Moore's Law in regards to computing power will render stealth much less relevant over time.
Like the C-130 and the B-52, some designs are simply classic and require only evolutionary updating for their mission.