Is the Modified Early Warning Score (MEWS) Superior to Physician & Nurse Judgement?

We've all noted the Modified Early Warning Score in our daily workflows.
But is this MEWS better than our clinical gestalt at predicting clinical illness?
That's what these authors sought to sort out.
Hat tip to the authors. Read these data for yourself.
They performed a multicenter, prospective observational trial of all patients who presented to the ED via EMS.
Let's just say that we kicked the butt of the Modified Early Warning Score.
800 patients were analyzed. 
As with any test or score, we need to plot out how specific and sensitive it is.
To simply this, we use the area under the receiver operating curve.
ED nurses kicked the butt of the MEWS Score with an AUC of 0.809 vs. 0.73
Physicians also kicked the butt of the MEWS with an AUC of 0.848 vs. 0.73.
For the sake of statistical context, an AUC of 0.7 to 0.8 is considered acceptable.
This means that the MEWS is just acceptable.
Meanwhile, both nurses and physicians fall into the excellent category with an AUC between 0.8 to 0.9.
It's also pretty neat that both physicians and nurses has equal skills sets in recognizing patients who are going to crash.
Do we really need all these pop-ups on our EMR or just good clinicians?