The sensors and computers to make this safer and not add thousands to the auto is only years old. I'm going to write less than a decade.
However you have a new generation that grew up with smart phones and won't know about this.
I see no mystery here.
The answer is, "Yes," but there is a caveat: Is the circumstance just one car or is there a consistent danger from either the hardware, software of the design that makes it possible or likely that the Cruise-Control will cause an accident. The cause for this discussion was tragic and most of the comments did not help me much at all. While unintended acceleration via operator error is the most common, one comment brought out in comments is important: the auto industry and NHTSA have not yet demanded that cruise control not automatically employ or deploy "all out acceleraton." This is part of a hard manual hold on the resume or 'accel' feature on the (+) side. There is no reason that feature should exist, because it is so dangerous that any benefit from its existence is overwhelmingly offset by the danger to the drive and his occupants and other drivers and property that such ability should be eliminated. If such speed is required for any reason then manual acceleration should be employed or used.
A second useful manufacturers design change should be that when a memory is in cruise control and more than five minutes elapses, the memory should be erased. If 5 minutes is too short, there should be some sort of arbitrary number that can lead to safety that should be selected and imposed. It seems to me that interruptions of longer than 5 minutes, when conditions can change, is plenty long enough. Another change is distance and elevation. If either distance of several miles has been driven with no cruise control used, the memory of the last speed should be erased. (Long distances driven without re-engaging cruise control is a sign of inattentiveness or sleepiness. It also shows that conditions are not right for cruise control use over an extended distance. Again, the law should require that cruise control be cancelled including last speed used when the brake was applied.
The final requirement for cruise control is in design location. The cruise-control, in particular, should not have its engagement switch lined up with any other driving or luxury aid, nor have them in close proximity to one another and they should not exist in the same plane of motion to activate them. Certain German manufacturers have not placed cruise-control buttons on the steering wheel, but instead have the cruise control functions all exist on a lever or a stalk, mounted on the steering column (sometimes incorrectly referred to as the steering wheel.). Because of this design, a person wanting to make a right lane change or a right turn, or move to the right in parking, could accidentally engage the "resume" function or the "Accel" function on the cruise control. If they are at all inattentive for any reason, or they have turned the steering wheel for driving purposes to a different that straight ahead angle, a reach for the directional signal (right) could easily result in inadvertent application of the cruise-control. If that happened while the memory of the last cruise control speed was still in the "RAM" or access memory of the cruise control system, present technology does not have a "smart" application of the cruise control. One could be doing 20 miles per hour and when the lever for cruise control is inadvertently struck while ostensibly reaching for the directional signal (right), the cruise control would engage and would do so at, "full throttle." This is just plain unsafe. Perhaps at the end of a long bottle neck of two lanes, where 4 lanes are suddenly available, and traffic opens up, the application of full throttle might be OK. But in real world driving, when inadvertence is part of the reason that the cruise-control became engaged, the traffic conditions most certainly would be too dangerous to allow for this type of acceleration. Because of miniaturization, chips, circuit boards and various interval functionalities, it would not cost much money at all to ensure that while both intentional and unintentional re-activation of the cruise-control was deployed, acceleration would be limited to a gradual increase to the speed stored in the memory. An even better solution would be for time and/or distance (which ever comes first) to erase the speed held in the memory. This would require the user to reactivate the cruise control and then come to a speed and hit the "set" button once again. Set would be plus or minus as traffic conditions allowed.
And additional fail safe for cruise control should include something that finally exists in 2019 and for a couple of years: Distronic cruise control where distance and speed of the car in front of the driver's car is monitored by cameras and radar and feed back and measurement of the speed and distance from one another of both cars. This should be kept constant by the electronic part of the cruise control system.
In some cases the Distronic feature allows for autonomy where the car speeds up or slows down with the car immediately in front of the driver, even if the car in front comes to a stop and then a short time later, starts up again. The car behind with Distronic monitors the car ahead and mimics that car's behavior, always maintaining a safe distance. (Hence the development of the name of Distronic with 'Distance and Electronic').
Lastly, sensors in the car should monitor traffic in the immediate vicinity of the driver's car. Should the Cruise-Control be activated with a memory function in this situation, the braking safety function should immediately be deployed to stop the use of the cruise control altogether and also cause deployment of the brakes at about the 60-70% level. Thus even if a failed cruise control bit of electronics cause a sudden acceleration, the braking system would come on to attempt to prevent that error-caused or failure-caused cruise-control- acceleration.
Finally, as suggested before, relocating all cruise control levers away from the steering column, if there is any other control present in the same plane and within inches of another lever that requires the same motion to activate, would prevent what had been, "accidents-waiting-to-happen." Why the manufacturers have seen this and haven't employed these easy remedies long before now is a mystery to me (unless it is money-related...oh yes. It is always money related!).
Your thoughts?

Chowhound
Comic Vine
GameFAQs
GameSpot
Giant Bomb
TechRepublic