Marut was an aerodynamically sweet and the machine was somewhat survivable for its day (can't say about the cockpit). Probably somebody may be able to jam a Kaveri into a Marut but that will most probably destroy its aerodynamics and then the people trying it will have to change a lot more than just the engines. In any case the world has moved on even in terms of CAS. Not worth the trouble, I guess.
In any case LCA is good enough for most of the CAS, as it is likely to be executed in future. See we have to get out of this desire to visualize an A-10 or a Su-25 whenever we think of CAS. These two aircrafts were very well built for the purpose but even these machines had to adapt to the threat environment and the advancements in the sensors. Within Afghanistan itself saw some of the Su-25 bombing run tactics change, which ultimately had them start out at almost the top of their ceilings and use dumb bombs with sights only. And for a fairly acceptable effect. Su-25 also delivered LGBs from varying ranges, again to good effect.
My idea of a CAS today is to make the thing in a way that the pilot gets some distance more before he has to eject, if he is hit. Because that hit is a forgone conclusion. Between a dual-color sensor and IR counter measures the balance of capability tilts towards the dual-color sensors. And its only going to be more and more difficult. Now whether you achieve this survivability by modifying the cockpit &/or the engine placements &/or armour, is a matter for engineers & ASQ writers, to decide.
When we think of CAS we immediately refer back in our minds to the youtube videos. But they are not a representative way of how we should be approaching the issue today and in future. Here this is what Sqd. Ldr. Ajay Ahuja said before he bailed out : Ahuja gave a radio call – "Hercules, something has hit my plane, possibility of missile hit cannot be ruled out, I am ejecting over...(location)."
Notice he had time only for that much and notice also the follow on results. Would you send your pilots into a like situation? Because that is exactly how much time a CAS pilot will have in his hands if the CAS mission goes haywire.
Today you want to run a CAS in an uncertain situation in a new battlefield, you can still rely on a lot of sensors to lead you to the launch point and do the final kill for you - between modern versions of GPS+INS, Terrain following radars, targeting pods and good availability of LOROP there is very little that the ground soldier would find as a big surprise in his fight, demanding risky CAS missions. The larger number of smaller surprises that the ground forces will find today will be missed out even in after a successful CAS.
In any case our Jags and Mig-27s are going to be around till we get the LCA to do like missions with similar survivability capabilities.