Specifically attenuverters with a switch, like the ones on Shades (and quite a few other modules.)
When I flip that switch from normal attenuation to attenuversion, it’s because I want to invert my signal. But now I’ve lost half of the knob’s rotation range because zero is in the center … so if I want to adjust my inverted signal I can only use 50% of the pot to do it. This makes fine adjustments more difficult. It seems to me it would be much more useful to have the switch activate an inverter, either pre- or post-attenuation, and that way you could still use the entire sweep of the knob to adjust the amplitude of the inverted signal. That makes much more sense to me. But I haven’t seen anybody do it that way. Why?? Is there some electronic-engineering reason why Attenuverters Must Be Like This? I know it’s trivial but it popped into my head a few days ago and it’s been bothering me ever since.
With a 3-position switch, it would be possible to design a circuit with 3 configurations: attenuation, attenuation of the inverted signal, attenuverter.
Waiting for a long facebook post about shades, why it is discontinued and that it will eventually all makes sense when the next modules come 
The SSF Mixmode has straight inversion switches on each of the unipolar channels, it’s sometimes useful sure but I usually reach for Shades (or Blinds) before that.
sorry - I don’t get it - did you read somewhere that shades is discontinued? because on the MI web-page it is still not under “discontinued modules”… would be a pity…
I think it’s a riff on the recent purge of modules, being replaced with newer, more refined implementations of the original idea.
Blinds kinda is a more refined version of Shades, but I think they’re different enough to both have their places, and I can’t see either being replaced soon.
In my experience this is because the idea was that many CV modulation sources were bi-polar, and so this configuration of attenuverter was a simple way to scale a CV that was already swinging across the zero-point - just more like a “shift” function. At least this was how we often thought of them in Serge-world. I can see that if one is using unipolar CVs what you describe would be very useful, but one can also just patch a simple inverter either before or after the “normal” attenuator - a little panel of home-brew inverters would not be hard to build, I think…