RBE971 V12.1.11.15: Intel AX211 client severe latency on main 5 GHz SSID, fine on IoT SSID

Thanks for the reply, and for the testing you're doing with the various adapters.

To your questions:

• Intel driver: 24.70.0.3, dated 7/29/2026, so we're on the same current release. It hasn't changed since before the problem started.
• Factory reset: Yes, full reset on Sept 29. I tested on the default configuration (default SSID, nothing restored) before touching anything: laptop beside the router on the router's 5 GHz radio, channel 40 (Auto's choice), ping avg 329 ms, max 1537 ms, 1 loss out of 60. Same laptop through a satellite in the same session: avg 6 ms. Then I restored the backup; no difference either way.
• Satellites: Two RBS (RBE970), both on other floors and on the opposite side of the house from the router, well over 30 ft away. Both show "Good" backhaul on 5 GHz + 6 GHz. Home is roughly 3,500 sq ft over three levels. The router is on the lowest level, and each satellite is on a separate level.

On the adapter question: I agree the AX2xx cards can be finicky, and that was my first theory too, but the results say it isn't the client here. A second laptop with a Qualcomm QCA61x4A (Wi-Fi 5) gets the same erratic latency on the router's 5 GHz radio (avg 301 ms, max 2140 ms, at 100% signal), and a Sony TV on the router's 5 GHz started buffering the same week. Meanwhile, every one of those clients, including the AX211, is clean through either satellite on the same SSID and channel (avg 5–6 ms, 990 Mbps). The AX211 also runs clean on a Linksys Wi-Fi 6E router placed in the Orbi's exact spot, on channel 36. So, a USB adapter would just be a fourth client showing the same thing; I'm happy to run it if you think it adds something, but three different radios from three vendors already agree.

The pattern is one radio, in one unit, on specific channels: the RBE971's 5 GHz client radio is bad on 36 and 40 and fine on 44, while the satellites' 5 GHz radios are fine on all of them. Manual channel 44 is my workaround right now, and it works well (1,084 Mbps, 3 ms pings). But channel 40 got measurably worse between Sept 23 and Sept 29 (from ~150–430 ms to avg 1117 ms with 16% loss), and it survived both the factory reset and a rollback to 12.1.10.12, so I'm treating 44 as a stopgap rather than a fix.

Support's position so far (case 49892824) is that since 44 works, the radio is fine and it's "channel conditions." The Linksys-in-the-same-spot test is what I'd point to on that; conditions don't discriminate by brand. If you can flag the case for a second look, I'd appreciate it. I'll keep the thread updated with whatever comes back.