That was probably this thread - Good Clock Modulators? - #8 by auretvh - on modulating clocks in general, where @auretvh also proposed using an inverted square wave from an LFO and modulating the pulse width. The gimmick being that it’s now the rising edge of the pulse instead of the falling edge that is … uh … swung, yet the period remains the same.
I suppose it would be a trivial thing to have the sequencer do the PWM on a per-step basis if there can be CV output dedicated to just that, such as with the Fundamental SEQ3 module.
Now that I think more on it, there might be a chance for “fence post” errors as to exactly when the pulse is modified. Unless it doesn’t matter.