Here's my take on utilizing random within pitch sequencing.
While it's great to have a source of random cv that can loop within a boundary of 8 or 16 steps (like marbles or turing machine) the actual listenability of those sequences can be hit or miss (and you don't want to take the potential 'miss' when you are jamming or god-forbid performing). I personally use an O_c for my turing machine purposes, and instead of sending the random sequence straight to an oscillator, I use it to address inputs on my sequential switch (plenty of sequential switches can do this), which all have simple cv sequences coming from my sq-1 or digitakt via midi-cv. This way there is still an element of unpredictability, but I am able to prepare sequences beforehand (on the sq-1 and digitakt) that will sound good when spliced together.
I find that this technique works really well for acid bass lines, but depending on the style of music you are trying to make, you might want to add your sequences together instead of swap between them. In my mind this is a good way to make long and virtuosic ambient stuff, where you have multiple sequencers clocked at different speeds that are then summed with a precision adder.