I would like the learn the Russian Shahed's price, success rate against defended stuff and accuracy. 40 kg is definitely not enough if accuracy is not superb. The Mk-81 was withdrawn from the US service because its effects were found to be not enough. Even the JDAM kit for it was canceled. 125 kg class bombs only returned to the US service in the form of SDBs, a purpose built gliding bomb with a very sophisticated guidance package. The recent ones, in some conditions, achieve a sub-1 m CEP. If the Shahed's accuracy is like 10 meters with that 40 kg payload, then it would be too weak for a lot of things. And if adding some more accurate and jamming-resistant navigation bumps their price to 100k or so, why not just build more CJ-10s then? It would terrain hug at Mach 0.8 and deliver a 500 kg warhead within meters. I would take that over 6-8 Shaheds anytime.
I think Shahed-type ammo is useful within rocket artillery distances. I mean sub-300 km by that. At that distance communications are easy, you don't need to fly for hours over deep enemy territory, etc...
Here is the Harop. The Harpy was among the original loitering munitions and the Harop succeeded it. Man-in-the-loop, good control software, ESM and IR, etc... Limited to 200 km by communications. Costs millions. Israel's concept for kamikaze drones was a bit different. Rather than cheap cruise missiles, they are more like loitering and situation-aware weapons that can be deployed to places where a threat can emerge. No deep strike or facility destruction purposes were considered.
Suicide drones aren't meant to take out small, mobile targets. They're there to hit enemy infrastructure.
The rocket force has a plethora of weapons for high end targets already. Having stroke piston powered drones allow them to get a larger piece of the strategic bombing action, before the air force can step in after SEAD is done.
That's all I'd see such platforms as, they're not some wonder weapon that will destroy a ton of enemy high or mid value platforms. But they will expose enemy air defenses to SEAD and will help China gain a powerful psychological advantage right out of the gate against the aggressor parties.
The obsession of cost - benefit trade is also misunderstood. More important than cost trades is production capacity.
War isn't won by not wasting more paper currency than the enemy. Otherwise you'd argue that cheap Tomahawk should simply not be intercepted by HQ-9s because the interceptor is likely more expensive? No, what's important is how fast a country's industry and replenish those platforms.
What is important to look at isnt if China's drones "cost" 100 000$ or 20 000$, what should be looked at is how much capacity to build such drones can exist, when factories start to be converted.
I would argue that China can afford spending way more money on every trade (around up to max 200-400% "unfavorable money trades").
Not only is the US economy around 30% smaller than China's, but as even Americans would know, a much larger share of China's economy is production and product development, vs banking and service sectors in America. The latter don't help build missiles/interceptors faster.
US is also going against tyranny of distance, adding even more extra costs to them.
Finally, China would be fighting for the freedom of their homeland, while US soldiers might be motivated by a vague sense of crusading against communism or something, but the point is that US soldiers would not be as motivated on the offensive as Chinese soldiers would be, fighting over their literal homes. So the cost of maintaining the morale and overall cohesion of the force would be more expensive for America as well.
So I would believe that unfavorable platform to platform trade cost is a red herring. China can easily go into seemingly really bad cost trades and still successfully repel an invasion.
Even if China's drone should end up expensive on paper, the industries will simply be forced to make it. Cut their profit margins and/or make people buy it with war bonds.