"DJI has only ever made products for civilian use; they are not designed for military applications,"

'cause if it was for military use, paid for by a governemnt (any govm't), the price would go up tenfold (at least, likely quite a bit more) for the same product...

Currently, drones designed for the military cost, what? Hundreds of thousands each, maybe millions?

If it turns out that you can make ten or twenty or more consumer-grade devices that are worse but still usable, for the same price, that starts to open up worrying new tactics and strategies. What do you do if your position get attacked by a thousand of these? What do you do if the opposing army deploys a million of them all across the front? What if they can replace lost drones at a pace of ten thousands a day? Sure, they are so crap that half of them just kill themselves, but the order of magnitude remains.

If I were a military guy, I'd start having some pointed meetings.

Combined with the research on drone swarms and control...

No need to buy a tank or a plane. Just swarm in and blanket the area.

[beware, the icon are swarms of drones; with jammers, of course]

Exactly. The cost of one fighter plane buys, what, a hundred thousands consumer-grade drones? Okay, maybe twenty thousands, once you've glued a weapon or a chunk of high explosive to each of them. They are crap, but after enough zeroes, quantity acquires a quality of its own.

There is a lot of bloat in military hardware admittedly, but it's not as simple as buying a a ton of consumer grade devices to replace the military ones.

Firstly, at the moment one operator can only fly one drone at a time. Yes if it gets shot down they could immediately launch another, but thats not particularly effective since the new drone still needs time to fly to where the old drone was to continue its mission, etc. But there are no actual swarm technologies currently flying. There are plenty of research projects exploring this, but there is no actual current way to fly multiple drones in a swarm in a military environment in a useful manner (the drone "firework display" type of swarm is not useful in the slightest in a military application).

Secondly, as has been found in Ukraine, tracking of civilian drone signals back to their operators is relatively easy. This means in Ukraine most of the time, the drones are flown very fast, on a recon mission, and then the operators haul ass to avoid retaliatory artillery fire. Military drones have encrypted and hardened comms to prevent this. In addition, the technologies exist now to disable civilian drones (jamming) at considerable distances. This so far hasnt been used in Ukraine, primarily because neither side seems to have this technology in place (surprisingly), however, the technology exists and is already in place in various airports and protecting military installations in the west. I am surprised this hasnt been implemented at least partially yet, but maybe both sides like having the recon ability? Again military drones are hardened against this sort of jamming technology.

Thirdly, civilian drones are only particularly effective at recon. They cant carry a heavy enough payload to carry a weapon that can deal a great deal of damage (in a military context), perhaps a single grenade, whereas military drones are usually much larger and designed to carry much larger weapons, up to and including multiple missiles. Just ask Afghanistan how effective those missiles can be at destroying things.

And on the recon side, a civilian drone usually has just a single camera. Military drones, usually carry various grades of optics (infra red, higher quality optics, etc.), can fly higher and with a greater field of view, and so are much more effective at observing without being seen.

The civilian drones have done wonders for Ukraine so far in the war, but as both sides start developing anti drone weaponry and technology, you will see the larger military drones having a much larger effect, as it did in the early days of the war when it was the large Ukrainian military drones, who effectively wiped out tank columns, with well placed missile strikes. That's not something the civilian drones will ever be able to do...

You might want to have a look at DraganFly and Drone Delivery Canada. They both have heavy lift drones with long range capabilities. The range is even further if the drone is on a one way trip.

Yes and no.

Yes, it opens up new tactics. However, if the devices are small then they also carry a limited payload. Non military versions also tend to offer no way of hiding from radio location, which means that a functioning army with radio location equipment can tell the artillery people where the operator is to drop a 155mm artillery shell on them.

If you deploy a million of them then this causes quite serious problems with charging, reloading and supplying them. Personally, if I was the opposition i'd simply wait until you'd deployed the lot, and then promptly blow away the local power infrastructure your using for charging, retreat a hundred miles and blow the bridges on roads between the drone force and it's resupply to complicate you resupplying. At the point you run out of either juice to charge with or munitions then they become effectively useless, and you could advance back over the top of your collection of useless drones. The military tend to say amateurs study tactics; professionals study logistics" etc.

Logistics is always the biggest headache of any advancing army; see the utter chaos the Russian army in Ukraine caused by a simple strategy of "blow up ammunition depots".

However the issue of large numbers of cheapish things winning against a small number of expensive things is well known; in the early 19th century the "Jeune École" school of thought was that explosive shells would doom [wooden] battleships. Then swarms of fast torpedo boats would doom battleships. It only actually happened a good century later when swarms of torpedo carrying aircraft did it; fundamentally economic power expressed by building a large number of cheap, reliable and deployable weapons that are effective against expensive stuff wins wars.

The Australian Youtuber I Did A Thing recently strapped an MP5 to one of these exact dogs and it had a lot of trouble a) staying upright on anything but paved or flat surfaces due to the altered COG, b) was overly complex to aim and c) crucially fell over from the recoil.

Robot dogs with freakin' guns strapped to them make for a great press piece at an expo, but I don't see them replacing us meatbags quite yet.

Do you buy 1x$1,000,000 super duper killing machine.

Or do you buy 1000x$1,000 killing machines ?

Noting new here. Just look at what happens when humans are aplenty and cheap.

We need to define quantum drone features. 1 drone is pretty much transparent to a high speed (~short wavelength) weapon whereas 1000 well placed slow ones might be seen as a solid to the same incoming weapon.