Is it normal for the burst rate to slow down when you stop down the aperture?

When you are looking at time periods of fractions of a second, the human standard of “near instantaneous” is the wrong unit of measure. Anything that physically requires as much movement as an aperture mechanism likely requires many milliseconds to accomplish. Those milliseconds might not seem like much to you, but they can have huge impacts in situations where we are slicing seconds into fractions. What makes matters worse is that there is clearly something in Canon’s aperture mechanism communication protocol that includes confirmation of step-down. If you ever have an aperture mechanism fail, you will find that you can continue to use that lens wide open, but you will get an error when you try to step down and shoot. That error won’t appear until after you press the shutter button all the way down. How does the camera know that the aperture didn’t close as requested? Clearly there is some verification being performed prior to taking the shot. All of this takes time, too. This all happens instantaneously by human standards, but none of this is even close to zero when we look at fractions of seconds.

I didn’t verify in all my DSLR manuals, but the fine print in the 5D3 manual confirms that the maximum fps is only achievable at wide open aperture. It also included two other factors that could slow the frame rate that I found difficult to comprehend, “subject conditions” and “brightness”. I suspect those can affect the processing rate of the image, itself. What you are seeing sounds normal to me.