I've done test strips both ways, continuous using a card to block the light, and with on-off increments.
Currently I use the on-off method because the timer I now use is programed for that.
Both methods work, I've not noticed a difference, but I've never tried to make a comprehensive comparison either.
I think that in practice any differences get taken care of by the total workflow. That is, you make a test strip, determine a best choice exposure, make a print and then tweak things until it's right, by some combination of exposure adjustment and dodging and burning. Step one is to pick a method and use it.
FWIW2, yeah, I agree regarding the "Fred Picker" method.