My suggestion for this is to have 1 primary/summary question that covers the behaviors that you expect the agent to demonstrate. In your example - the question group would be Customer Experience - with a 10% weight. The first question could be an all inclusive "Did agent demonstrate all behaviors to support a positive customer experience? (tone, professionalism, empathy, call control, rate of speech, language/communication skills)". I also suggest separating professionalism from empathy - as those are different. And while you always expect professionalism - all calls don't necessarily require demonstrating empathy. Also control of the call has specific things the agent needs to do - while rate of speech is yet another behavior.
This summary question would have the desired scoring value associated with the “yes” - contributing to 10% of the overall score. If the response is "Yes" - then no need for the evaluator to the respond to the additional questions about each of the behaviors.
You can design it so that the additional questions only appear if the answer to the first one is “No” . So when you add that second question “Appropriate voice tone” set it to “display conditionally” when response to the first question is “No” Each of those additional questions can be set with a 0 value so there is no impact to the score.
I’ve attached several screen shots as examples.