How obvious is it that the plane has entered non-icing conditions?
The problem isn't knowing whether or not you're in possible icing conditions if you think to ask the question, but rather reliably remembering to evaluate it every single time you enter and exit possible icing conditions.
* SAT - static air temp (air temp before the ram rise).
It should be a caution light or something at least.
But ATPs don't have any trouble evaluating "am I in potential icing conditions?" as they've been doing it for one to many thousands of hours previously.
This kind of mistake has been made so often and lives have been paid that I find it crazy that it's still being proposed.
If it were just a temperature thing, you'd think it could be automated, but I don't think that's really the main thing they're dealing with here.
In theory this means switching the system off when you exit the clouds.
This is usually pretty noticeable, but maybe less so at night, since you might be breaking out into a pocket or a clear layer between other layers, without visibly seeing much of anything outside.
I would hesitate to comment on the feasibility of this beyond what the interviewed persons have said, precisely because they're not clarifying (to the readers in any case) what the actual thresholds here are. And the interviewees don't seem convinced that this is a reasonable/safe requirement.
But easy to forget to check the instrumentation.