IANARS, but as far as I can tell, you would still have limited terminal guidance at the cost of significant payload.
A better investment is usually going to be with launch precision and boost phase navigation. For every dollar spent on improving the precision of the course during boost you're gonna get the equivalent of hundreds of dollars spent on terminal guidance with this sort of projectile, as far as I can tell. It seems backwards, but seriously, the challenge of steering something at 9 mach without destroying it is pretty significant, especially considering the type of buffeting you'd have. If you could increase the accuracy of an Iraqi SCUD by more then 10% through terminal guidance, I would be amazed, but again, IANARS, so someone else here might have a better answer.
The way I figure it, you need to start with the CEP of the existing round, then design a system that will maneuver within that circle.
The CEP for the SCUD A was over two miles. It is travelling at one mile per second coming in. You'd need to write a navigation system that would deal with the thickening air during the final few seconds and give appropraite control impulses to avoid overstressing the mechanism. You might be able to improve the SCUD A because of the relatively low speed and high CEP, but nobody uses the SCUD A anymore, it was withdrawn from service in the late 70s.
Take the SCUD D, on the other hand. Coming in at 9-10 mach, it's covering two miles a second at a 45 degree or so angle... but already has a documented CEP of 50 meters (according to some online sources, I don't know the actual numbers). Terminal guidance will give you little return, AND it's harder to implement. Not a winning combination, especially considering the already limited payload size.
Have I mentioned how well suited cruise missiles are yet for low cost, off the shelf, super high precision munitions yet?