Washington — David Harsanyi’s 29 September 2026 column in the Washington Examiner needed one military sentence to close a political argument: China is not an American peer. The sentence he used was this: recent fighting in Venezuela and Iran “have proven that Chinese-made air defense and missile systems are no match for advanced America’s stealth warfare capabilities.”
RNA’s 30 September recap went further and treated the same theatres as a factory verdict: Chinese weapons “failed again” in Iran after Venezuela and after India’s Operation Sindoor; Iran’s “Chinese-origin HQ-9B crumbled against F-35 stealth barrages.” That is the claim that has to be taken apart. Not whether the United States can fly in the Caribbean or over the Gulf. Whether those two campaigns measured the PLA’s air-defence force.
They did not. They measured export batteries, mixed with Russian shooters, in someone else’s command net.
1. What Was Actually in Those Two Skies
Venezuela’s long-range layer, in the open after-action file, was Russian: S-300VM and Buk-M2E. The Chinese piece that Western reporting named most often was not a HQ-9 battery at all. It was the JY-27A metre-wave early-warning radar, sold as an anti-stealth cueing set. BW Businessworld’s January reconstruction put the problem where it belongs: detection is not a kill. If the radar does not talk cleanly to a Russian missile, or if the missile battery is jammed, sabotaged or out of interceptors, a “Chinese radar failure” is an integration story. It is not a scorecard for a PLA HQ-9 brigade outside Shanghai.
Iran is messier, which is why it is a worse place to run the same headline. Open reporting since early 2026 has described a possible HQ-9B export sale or integration alongside S-300PMU-2, Bavar-373, Khordad-15 and short-range Russian and Iranian guns. WION and Defence Security Asia treated HQ-9B as a reported layer, not as a photographed, battery-by-battery kill list. RNA then converted that reported layer into a collapsed system and a fake-claims indictment of the whole Chinese defence industry. That leap is the error. A strike package that reaches Tehran can mean empty magazines, broken cueing, too few batteries, or a radar that never saw the raid. It does not, by itself, mean the domestic HQ-9 family would behave the same way under PLA doctrine, PLA electronic-war support and PLA reload.
2. Export Catalogue Is Not the PLA Stack
They are not required to be the same software load, the same radar, or the same training syllabusStealth beat ChinaStealth beat a thin, mixed, customer-operated net that also included Russian long-range missiles.
| What Western copy collapses | What the open file keeps separate |
|---|---|
| “Chinese air defence” | Venezuela: JY-27A radar. Iran: reported HQ-9B export. Export lot: FK-2000, FD-2000, KS-1A, FM-90 (Ashgabat, 27 Sep 2026). PLA inventory uses HQ-9 / HQ-16 / HQ-17 and a different command net. |
| One lost campaign means the industry is “rotten” | CASIC builds the export truck and the domestic battery. Same factory does not mean same software, radar, or training syllabus. |
| “Stealth beat China” | Stealth beat a thin, mixed, customer-operated net that also included Russian long-range missiles. The Examiner sentence never names a PLA unit. |
The FK-2000 point is the cleanest illustration, because it is already public. China Daily’s 28 September note on the Ashgabat parade listed Turkmenistan as the fifth FK-2000 operator after the UAE, Chad, Ethiopia and Sudan. The same week’s Chinese-language technical file still treats FK-2000 as an export gun-missile truck, not as a PLA inventory item. Using a Gulf or African FK-2000 shoot-down — or a Venezuelan radar that never handed off to an S-300 — to grade an Eastern Theatre HQ-9 battalion is the same category mistake as calling Type 076 a fourth carrier because the deck is long.
3. The Western File Already Contradicts Itself
Four days before Harsanyi’s column, CNN reported U.S. assessments that Chinese geospatial data had improved Iranian timing against ships in Hormuz and had been a factor in a July strike on a U.S. base in Jordan. The same piece said China had also sent air-defence assistance. You cannot, in one week, file China as the reason Iranian drones are landing more accurately and, in the next column, file “Chinese air defence” as a proven toy against stealth. Both sentences can be partly true only if they are about different objects: commercial or grey-zone imagery on the one hand, a handful of exported launchers on the other. Neither object is the PLA integrated air-defence system on the mainland.
A correction does not claim that HQ-9B export batteries in Iran performed well. The public record does not support that claim either. It does not claim JY-27A saw F-35s in Caracas and was ignored. It claims this: peer is a force-on-force word. Two customer wars do not complete that test.
Bottom line
Recalibrate Harsanyi’s sentence until it names the product. Venezuela tested a Chinese radar in a Russian-shooter net. Iran, if HQ-9B was even on the firing line in useful numbers, tested an export long-range battery inside a battered national grid. The PLA stack was not the defendant. RNA’s “crumbled against F-35” line should be read as advocacy copy, not as a battle-damage assessment of a Chinese theatre air-defence brigade. When a Western outlet wants to argue China is not a peer, it will need a comparison that uses PLA organisation, PLA density and PLA reload — or it should drop the word proven.
Sources: David Harsanyi, Washington Examiner, 29 September 2026; RNA, 30 September 2026; BW Businessworld on JY-27A and Venezuelan S-300/Buk cueing, January 2026; WION and Defence Security Asia on reported Iranian HQ-9B integration, 2026; CNN, 25 September 2026, on Chinese geospatial support to Iranian targeting; China Daily, 28 September 2026, on FK-2000 as an export system. No claim is made here about classified PLA radar performance.