From this preliminary study, we conclude that speech synthesis over the telephone is not a solved problem for an increasingly important part of our population and that research efforts need to be concentrated on changing this.
From this preliminary study, we can already see that there is a decrease for all types of speech over the telephone channel from 100% comprehension at ages 20-29 to just under 60% at ages 80-89. And, clearly, the synthetic speech we used is less understandable than natural speech. These findings imply that more natural synthesis quality is necessary to improve understanding.