Oh, I didn't expect any more comments on this subject at all, therefore i didn't regularily watch it. So please excuse my delayed answer.
The actual answer is "no" anyway. I did not make any quantifying tests at all. I was busy and eventually happy with getting it to run the way it was supposed to. I developed and tested it on genuine 486 and 586 machines under MS-DOS where it performed "sufficiently" fast, that's all i can say.
I'm glad to see that it seems to work for you too. There is often the problem of forgetting "special" conditions during the tests, as was the case with the earlier versions. Therefore it is always good for a program to be tested by someone who did not develop it.