Short answer: no. I just wrote a long diatribe attempting to explain why using my very limited knowledge of semiconductor physics, but realized I couldn't do so without making some substantial guesses.
So instead, I'll offer some (probably fallacious, but likely still correct in its conclusion) logical deduction -- If this was likely to be a good photovoltaic source, the inverse would also likely be true. That is, that good photovoltaic sources would emit heaps of light when a current was applied.
In short, LEDs are semi conductors optimized to have a very specific direct energy band gap which produces photons in the visible light spectrum. This causes the semiconductor to shed excess electrons as visible light energy.
Photodiodes (used for PV arrays) are similar, but optimized for a completely different goal. They attempt to maximize the size of their depletion region. I don't know that the energy band gap is as important, though I'd imagine the narrower the better. Either way, this is done to maximize the likelihood that an incident photon will excite one of the molecules in that region of the diode, thereby freeing up an electron and causing it to flow around the circuit.
FYI solar cells would not make good light sources since the most common ones (silicon based) are indirect band gap. For LEDs (and lasers) a direct band gap material is required.
You are correct about the depletion region needing to be large for solar cells, but narrower is not necessarily better. This is because if you make the band gap super narrow (say 0.5eV) and most of your energy comes in as much higher energy photons (1eV+) you are discarding all of the energy besides that 0.5eV from those photons. The band gap is tuned to get the highest responsivity at wavelengths where the most energy is coming in if it is tuned at all.
Nope! In very rare cases it can generate multiple electron hole pairs the photon if it has enough energy to do that. Almost all of the time however it will generate a phonon (aka heat) rather than a photon with the extra energy. A phonon is a wave in matter rather than electromagnetism (photon).
So instead, I'll offer some (probably fallacious, but likely still correct in its conclusion) logical deduction -- If this was likely to be a good photovoltaic source, the inverse would also likely be true. That is, that good photovoltaic sources would emit heaps of light when a current was applied.
In short, LEDs are semi conductors optimized to have a very specific direct energy band gap which produces photons in the visible light spectrum. This causes the semiconductor to shed excess electrons as visible light energy.
Photodiodes (used for PV arrays) are similar, but optimized for a completely different goal. They attempt to maximize the size of their depletion region. I don't know that the energy band gap is as important, though I'd imagine the narrower the better. Either way, this is done to maximize the likelihood that an incident photon will excite one of the molecules in that region of the diode, thereby freeing up an electron and causing it to flow around the circuit.