I may be about to say a great nonsense
I don't think so! SLRGear.com
produces Blur Index plots using DxO Labs
technology. For those unfamiliar with the SLRGear reviews if you click on the Blur Index plot to the right of the text review a browser window opens which allows you to see the plot change as you move the sliders for f number and, where appropriate, zoom.
Have a look at their review of the Canon EF 135mm f/2L USM
which, from the review, is "one of the sharpest and most distortion-free that Canon makes". At f/2 through f/16 this lens is pin sharp and only loses this sharpness at f/22 and f/32 due to diffraction. This lens costs around £650 in the UK if you shop around.
Compare this with the SLRGear review of the Canon EF 85mm f/1.8 USM
which pricewise, at £250, is more in the comfort zone. The Blur Index plot shows that this lens only gets near the EF 135mm f/2L by stopping down to f/5.6. Shoot at f/2 and the lens can't match the size of the camera sensor pixels.
Given that a teleconverter is going to soften an image and also, unless you override the camera, mostly work the lens at its biggest aperture it is my guess that, especially with a tripod, Tom's choice of a mirror lens is the correct one unless
you are already fortunate enough to own a pin sharp lens to put in front of it.
Gordon uses a different testing methodology so it would be interesting to hear his views on this. I wonder if CameraLabs has considered testing Nikon/Canon teleconverters? Difficult to do well as, regardless of the sharpness of the lens in front, teleconverter performance varies with the focal length of that primary lens.