The part that I don't buy into is the idea that currently buildable linear particle accelerators could generate thrust at an efficiency of better than ten millinewtons per Watt of electrical power--that is a full two orders of magnitude greater energy efficiency than current bleeding-edge ion thrusters. He seems to be assuming that the electric current fed into the particle accelerators will be converted into particle momentum with better than eighty percent efficiency, making negligable allowances for refrigeration of the superconducting magnets or magnetic energy lost to interaction with things other than the reaction mass or anything else for that matter.
Very good example of why I posted this here. I have no idea about that stuff, and couldn't've made
half that paragraph...
