Unless you are a well-experienced micro-electronics engineer, work for the camera manufacturer in question, AND have access to a motherboard manufacturing factory, it's unlikely anything other than some minor software tweaks for image processing can be done to the camera.
As mentioned before, the camera is mostly a computer with an optical sensor attached, and a mechanical shutter on DSLRs. With any computer, "throughput" (how fast it can process stuff) is constrained by the limits of electrical transmission speed (not an issue in a camera), transmission/reaction times of each chip and it's component micro-transistors, storage bus speed (connections to and from the memory card (it has to 'confirm' what it just recorded! or, at least, acknowledge it just received 14 bits of data)), and lastly, the 'reaction speed' of the individual senels in the sensor plate.
While we humans think an exposure at 1/5000 of a second is super fast, in terms of what speed a computer processes, we're talking 1 or 2 nano-seconds per instruction! (billionths of seconds). So 1/5000 sec is an ETERNITY for a computer. To make the computers faster, desktop and laptop processors have embedded data storage right inside the processor chip itself to eliminate the 'waiting' necessary to get the next instruction and data from 'storage' (a RAM chip or more somewhere on the motherboard). Even the microcoding embedded in the physics of the chip is a 'science' of its own to optimize data access times across multiple processors communicating across multiple buses to multiple storage locations...all inside a single chip smaller and thinner than the fingernail on your pinkie! Simply put, the internal workings of a computer processor are far beyond most people, even beyond many electrical engineers, as well. It's also exceedingly proprietary and "top secret" at the chip manufacturers such as Intel or Motorola. And who knows who made what chips in camera A...or camera B... or which revision of those chips were used since such and such date, etc...
So, unless you are a high level technical engineer at Canon, Nikon, Sony, or wherever, the likelihood of being able to make a software 'hack' of any kind is somewhere between positive zero and negative zero (yes, all computers can represent negative zero!)
And I'm just a retired computer geek...what do I know?