I know that this is a bit extreme, but a highly placed source has told me that depending how badly "they" want to identify a voice, tuned and formant processing can now be easily reversed with software that even has presets. The best way to disguise a voice (and vice a versa, done live to impersonate another, like photoshoping a person's diction) is by replacing it with a canned voice...speech to text and back to speech. That way it is untraceable. Spooks, eh?
I am lately feeling as if I am everywhere at once though no place in particular.