I'd agree that there are occasions where a "manned UAV" is a lot better than a UAV. The accident rates for UAVs are still higher than for manned aircraft, which offsets their cost-savings. I'm not entirely convinced by the advantage of a DA42 UAV. As you say, there are airspace restrictions on UAVs in most countries, so whereas you could just fly a manned DA42M to theatre, and once in theatre would have more diversion options as you could go to any airfield you like. Getting a UAV DA42 to theatre would be logistically more difficult, then you would be much more restricted in where you could divert the aircraft too partly because of airspace restrictions and partly because (presumably) you would need a pilot on the ground at the diversion to handle the landing (I doubt that could be done remotely).
Also, I'm not convinced that the DA42 is necessarily a good platform for conversion to UAV. Those long, low wings would be a bit vulnerable when trying to land one remotely in a cross-wind (even with a pilot controlling it locally), and detecting then handling an engine failure remotely would be a challenge. However, the surveillance and observation flights I've done mostly involve really chucking the aircraft about to stay in the right location and I can imagine it would be difficult to convert a DA42 to fly (say) low-level steep turns automatically. It would only be worth converting DA42 to UAV if you really had an operational requirement for very long endurance flights, all of which were conducted at mid-level (F150-F180).