I'm not much of an expert, but air warfare of manned fighter aircrafts seems to be limited to sorties against determined targets or patrols against possible dangers. If one combatant has a better system to activate his manned air power for strikes while the other must fly more patrols in order to know what's going on then even an outnumbered force of aircrafts can overwhelm a numerically superior enemy piecemeal.
That's a simplification to show my point that not numbers but effficiency of deployment count. The aforementioned smoke would be one of the methods an outnumbered US force would use to fight numerical Chinese forces. Especially aircraft carriers have a high sortie rate, can switch location and while not easy to spot, can have enough fakes around to make the depoyment of a strike system against carriers really dangerous (not counting the missiles, but the whole targeting complex).
It's my opinion that in modern warfare it's easier to fake an appearance and military supremacy is much dependant on the ability to fake and to see through these disguises. I'm not sure China has the same benchmark as the US and her allies in that field (altogether over 1 billion people) with lots of technically experienced people who had a lifetime on information technology jobs and can easily work from home in our interconnected world.
As a result I would step up US sensor capability in order to optimize resource deployment against an enemy with plenty of stuff locally available.
The other revolution in aerial affairs is permanence and availability. Strike and intelligence gathering drones are accessable on increasingly lower military unit level in staggering numbers with few countries other than the US having much knowledge about that same degree of deployment.
The increased aerial availability of unmanned platforms in my opinion is a force multiplier for the manned platforms due to increased damocles strike and surveillance capability. China may have reached some supremacy on the old playing field, but the US is rewriting the rules.
Having all new ships capable of launching UAV is of major importance for the new carriers that are far from dated with this new approach.
The F117 case was supposedly pulled off with HUMINT from a spy (a French general got blamed) and still seems to have been a rather difficult affair. I heard rumours that in some cases similar traps were set by the North Vietnamese against stealth aircrafts that have the characteristic of predictable directional reflection in a certain angle.
The whole debate about the F35 is about an idea that has ever since the F-104 starfighter been central to the American idea of a warplane. Small wings and high thrust with lots of gimmicks on board. It's possible that the F 35 will mature with improved thrust vectoring into an outstanding aircraft that even benefits from the small wingplane by increased capability to go unpredictable under less restrictions of airflow than other designs. Finally small things like the Libelle g-suit that works much faster than conventional designs and allows slightly higher g-forces will have a major impact on the capabilities a pilot has with his aircraft (imagine everyone becoming a g-monster). As far as I know the Libelle technology is not available to China and they have nothing equivalent.