Feature Request: Scale Denoising Strength Based On Mask Confidence Value

[jthaloreach]

Similar to the Power Scaling feature, could you add a feature that scales the denoising strength for a given tab based on the confidence level detected for the mask? This would help reduce undesirable consequences when an unitended object is identified as the object to be detailed.
I do understand that a potential suggestion may be along the lines of "Turn the mask confidence threshold up higher.". Unfortunately, this will not resolve my request, as some adetailer models I use require the threshold to be set low enough to be able to correctly identify any objects. I also understand this is in large part a shortcoming of the particular model I am using. Thank you for any help you are able to give.