@owen11 I hear you—this is genuinely one of the more frustrating gaps in Aqara’s current Matter strategy. You’ve already done the right thing by escalating through official channels, so let me unpack what’s actually happening technically and why your experience with the FP2 vs. W100 differs so dramatically.
The architectural split you’re hitting:
The FP2 is a Wi-Fi native device that connects directly to your router, not through any hub[“content_id”:8"]. When you “bridge” it to Matter via the M3, you’re asking the hub to proxy a Wi-Fi device’s state into Matter—which is a fundamentally different path than how Zigbee devices like the W100 work. Zigbee child devices live on the hub’s mesh and can be exposed with fuller cluster support because the hub owns their entire protocol stack[“content_id”:6"].
The M3’s Matter bridge functionality is designed primarily for Zigbee and IR devices it directly manages[“content_id”:6"]. Wi-Fi devices like the FP2 weren’t architected to be hub-dependent, so the Matter exposure path is more of a “best effort” translation layer rather than native cluster support. That explains your missing lux data and the brittle presence boolean—you’re essentially getting a synthetic Matter device rather than a spec-native implementation.
Why FP2 specifically lags:
From the reference materials, the FP2’s automation conditions in Aqara Home are actually quite rich (zone-based presence, absence delays, stillness detection)[“content_id”:3"][“content_id”:5"], but none of that granularity translates through the Matter bridge. The hub can only export what it can map to standard Matter clusters, and presence sensors aren’t fully standardized in Matter 1.0/1.1 the way switches or lights are.
Your instinct about the FP400 solving this is likely correct—if it follows the FP300’s pattern as a dual-protocol (Thread/Zigbee) device with native Matter-over-Thread, it would bypass the hub translation entirely and expose proper clusters directly to your Matter controller. That gets you the luminance (1024) and occupancy (1030) clusters you want without the M3 middleman.
On the broader ecosystem point:
You’re right that Home Assistant’s ESPHome dominance predates Matter, but there’s a silver lining—Matter’s adoption curve is actually accelerating because of that fragmentation. Aqara’s push toward Thread-native sensors (FP300, likely FP400) is partly a response to exactly this frustration. The W100’s solid Matter performance shows they can do it when the architecture aligns.
Since you’ve already ticketed this and it’s clearly a product feedback/architecture limitation rather than a supportable defect, your best leverage is continuing to surface this gap through the feedback channels you’ve found. The FP2’s Matter limitations are by design, not bug—and design changes need product-level prioritization.
This information comes from community discussions and user experience summaries and may vary by environment. Please verify carefully.[“content_id”:2"][“content_id”:6"][“content_id”:8"]
