Another speculative suggestion.
The max speed for flight with spoiler deployed, partially deployed or in transit is probably about the same. So there's no structural reason why you can't autodeploy from partial to full.
But flight slats-in could be all the way up to Vmo. If autodeploy was capable of initiating from zero slat, you'd have to consider both the case of the accelerated/high-g stall near Vmo and the failure cases causing inadvertent autodeploy at Vmo. The hazard involved would perhaps bias against auto-deploy from clean, especially if the characteristics were already reasonable without it.