Not exactly a great analogy as you cannot template the PC/Laptop industry to specialised Camera companies. The volume of scales are not the same and the CMOS sensor RD is very restricted between Canon and Sony.
Sensor fabrication defects also affect yield on CMOS sensors much more than they do on IC's. With an integrated circuit like a multi-core CPU, if a defect kills off a core or two, or even part of L2 or L3 cache, manufacturers can usually disable the unusable parts. That greatly reduces the loss due to imperfections in wafers, salvaging "damaged" parts for use as slower or less capable versions. This would apply to CPU's with large on-board caches, multi-core CPU's, GPU's, etc.
When it comes to CMOS sensors, a single defect tends to kill the whole sensor. When it comes to larger full-frame sensors, they start right out making FAR less efficient use of wafer area than smaller sensors, cpu's, gpu's, etc. You can't salvage a sensor etched onto defective wafer area, so they are pure loss. That greatly increases the price of FF sensors over smaller ones. Thats also a significant factor to why medium format cameras are so expensive (the least expensive of which are Pentax's 645D.) MF sensors use wafer space even less effectively, and defect loss is even higher (the greater the sensor area, the greater the chances of losing entire sensors to a single wafer defect.)
Comparing the larger format digital camera market to the laptop or even larger PC market in general is indeed a rather poor analogy. The circumstances and behavior of the two markets differ radically, on many more levels than just CMOS fabrication.