Well. nowadays I hate to say anything is impossible. Still packing electronics like that into a cannon shell would be a challenge.
Imagine you are in a rocket flying to a target. The motor fires for several minutes, up to the peak where you begin to fall. Now picture the same trajectory in a cannon shell. All the energy needed to get you to the top of the arc, the same amount as was applied by your rocket motor in the first example, is passed onto the shell in one big wham.
That is a bucket load of G-forces. Many, many. You would need to have very robust electronics to survive that sort of punishment.
OK so we get you to the apex of your trajectory. Then what? What mechanism would we use to guide your shell to its target? Fins? Rockets? These would also have to be robust and would detract from the explosive payload.
So we have some really big company build us this shell. Would it have much use? Well, yes.
Unlike aircraft, gunfire can hit and keep hitting almost forever. Further, we might be able to get shells onto a target faster than we can brief pilots and get them there with the right load.
On the other hand, shells cannot look for targets and hit them as they appear. Also shell are smaller than most every aerial bomb.
Before we spend big bucks on a project like I would like to see a lot of simulations and studies of the number of targets that appear how far inland and their vulnerability.