A Warpdrive. (Just because I am currently playing with GR as a side project.)
Actually serious, I believe the transmission of information between the frontiers of science and the public are broken. On one hand the university system is geared towards producing specialists, on the other hand the media is geared towards spectacular as orthogonal of important, due to a mixture of science journalists being the product of the university system and market forces. So one would need a reliable process to produce the one book to read about X.
In first approximation one could try to produce a four volume series "Frontiers in Physics", were the volumes are "Particle physics," "Atoms & Nuclei," "Solid state physics" and "Astronomy and Geophysics" and then distill it into a single volume. (As a illustration of Universities producing specialists, I could not name the four titles for chemistry, let alone linguistics.)