As for the first detection itself, I turn to the antenna principle to attempt to discuss it in a bit of detail. The antenna principle attempts to understand the absorption and emission of radiation on the basis of the behavior of radio antennas, which, although it may seem inapplicable to a quantum situation, actually is if you examine the mathematics closely.

The state function of an electron in a specific energy state is a function of x, y, z, and t in a way such that the absolute square of the function is everywhere time-independent, although the function itself varies with time according to the factor exp(iω_{1}t), ω_{1} being determined by the energy level. If the electron moves to a different orbit with a different different energy level, its state function is again of x, y, z, and t in like manner, except that the dependence on x, y, and z will in general be different, and that the time dependency will be exp(iω_{2}t), with ω_{2} being determined by the different energy of that state. But as before, the time dependency of the absolute square will be constant. The absolute square is associated with the probability that a process to find an electron at a given point would find it there, and thus can be thought of as equivalent to a classical charge density, and what this all says is that when the electron is definitely in one or the other state, its charge-density distribution is independent of time, which corresponds to a classical electrostatic situation. (Note that this is true despite the kinetic energy the electron possesses).

When, however, the electron is absorbing or emitting EM radiation, it will be, for a short time, in both states at once. Now if you add the two states, you will find that the absolute square is no longer time-independent, but varies at a frequency 2π(ω_{2} - ω_{1}), a frequency that corresponds to the frequency of the emitted or absorbed radiation. During this time, the charge probability density is varying with time, very much like the currents in a classical antenna. As such, this gives us a method of calculating the radiation pattern, i.e., the probability that the photon will be found in the various possible directions of flight. For the purposes of this experiment, therefore, the task is to devise a first detector that will have such available electron orbits so that when the photon encounters it, the electron that it excites will conform itself to its wave pattern during the transition to the higher energy level, in such a manner as to replicate that wave pattern as nearly as the loss of energy will permit.

Of course, when so doing, the photon must not lose all its energy to the transition process, but only a small part of it. Thus we must contrive that the electron when being energized by the photon not simply be ejected from the detector as in a phototube. Achieving this will likely require that the detector be carefully designed. It must absorb the photon, but then retransmit as much of it as possible on nearly the same frequency and with as near to the original wave pattern as possible. Properly designed, the detector would be able to respond to any wave pattern (within the wavelength range of interest) that were to impinge upon it, retaining that wave pattern and re-emitting the photon with slightly less energy to continue its journey. I am not able to describe the exact details of such a detector, but believe it can be done through a system of wide-ranging electron orbits made available through a correctly structured system of conduction bands that give us electrons that have no definite position in the transverse directions, as well as slight energy gaps that can be crossed using substantially less energy than in the incident photon. It may be that more than two such energy bands would have to be provided, and that the detector may have to have layers of differing material that would allow these processes to be carried out in sequence. It may be necessary, for example, to have an intermediate energy band whose energy is high enough to absorb most of the photon energy (so that the electron is not ejected from the material), from which the electron would then transition into the final energy band that would be just a little different than the incident energy. But the key to everything is that all this takes place in spatial phase, so that the original wave pattern is maintained. Doing all this may require cooling the material to a very low temperature to depopulate the higher energy levels and make them available to clearly receive the activated electron. With a first detector build on these general principles, the process should be possible.