They don't need to be that hard to develop. It's not like you need a return to base feature. Sort of like wanting to learn how to fly a plane, but not take off and land the plane. You don't need to have the full feature set of a DJI drone to make it a weapon. You also don't need to worry about the battery life of coming back, so you've essentially doubled your distance capabilities, so maybe a stronger radio. Do you need the video return signal too? I guess that would be some decent PR footage on the nightly news propaganda stations though, so might be worth the expense??? You also don't need to be burdened with the GeoFencing features of a DJI, but would be funny to see a bunch of attack drones all hover just outside a map boundary because the target has their base listed in the GeoFence library!
So while not duct-taping a GoPro (we'd use Gaff tape anyways), they could use bailing wire with a grenade or c4 bundle to attach it.
I think you're pretty confused about what most of these drones are doing.
DJI drones aren't being used as weapons platforms in the US...
They're being used for industry (agriculture, real estate, land surveillance, fire monitoring)
No one gives a shit that it's not difficult to make a flying grenade. They care that all the features you're in here mouthing off about as "not important" are actually important.