Because wildlife photographers are almost never 'too close' to their subject. If Canon made the Mark III so much faster, then they shouldn't have any problem pushing through 24 or even 28 megapixels and still have 14 FPS with a buffer around 100 shots. Why not bump the resolution just a little to create a buzz on the internet and maybe draw in some new customers instead of relying solely on the existing professional customer that you already know will upgrade. Canon develops products from a safe and conservative engineering perspective and not from a marketing perspective. Does it make business sense to rest on your laurels in a rapidly declining business segment? The professionals will upgrade because they probably already have a very tired Mark II with a really high shutter count and their business can more easily absorb the cost, but the not-quite-professional may see the incremental changes to the Mark III as not worth their hard earned $6500 and may choose to it that money toward a more reliable car. Sometimes you have to wonder if Canon makes its decisions based on the marketplace or based on their own engineering limitations. They always seem to be playing catch up rather than innovating.
Thanks for your answer, I think it makes a lot of sense, but I think to say that Canon isn't innovating isn't exactly fair. Even mark III's huge bump in the onboard processing power, high ISO range, weight save, AF improvement, new AF-on button, 16/20 FPS, and 1000+ buffer are all exciting features. To my eyes I haven't seen capabilities like these in any gear before.
With that said, I think mark III simply isn't the ideal bird/wildlife gear where pixel count is a high priority. At least it is not more so than X mark II, D5, or even A9, which all generally fall into the same pixel count bracket, give or take the extra 4 Mpx. But for it's purpose as a sports shooter, I still think it is heading towards the correct direction.
You mentioned marketing strategy of Canon. It's a good point and I agree that Canon isn't prioritizing its focus in wildlife photographer market with this camera. But for a good reason: the upcoming Olympics in Tokyo.
My personal opinion for bird/wildlife is that APS-C sensor cameras may strike a better balance between useful pixels counts and speed. For Full frame camera to deliver the same results, I think a high pixel body is a great choice if it can bring high enough FPS (8~10+), such as the D850, Sony A7R IV. I hope to see Canon's high megapixel R to match that ability, although I doubt 10 FPS on a 70+ Mpx is technically possible for Canon yet, but I'd love to be proven wrong.