I have a Magic Mouse 2 with usb-c to lightning port.
Not sure if assuming the external hardware conector is enough.
Also, I discovered in the log that in Mac OS 26 happens a WindowServer in the very last line:
2026-09-16 08:43:31.559077-0300 0x99b Info 0x0 105 0 WindowServer: (CoreUtils) [com.apple.bluetooth:CBDiscovery] Device found: CBDevice 07919989-33E5-B8B1-1405-ED427C237CC5, BDA <private>, Nm <private> , PID 0x0269 (?), PrNm Magic Mouse, VID 0x004C, VS 1, DsFl 0xA00000 < Connections Pairing >, DvF 0x40000024000 < ClassicPaired HIDBadBehavior Connectable >, DvT Mouse, RSSI -56, CnS 0x800020 < HID ACL >, SupS 0x20 < HID >, Freq 2.4, Battery M -64%, ClkH L NoiseManagement R NoiseManagement, FV '1.9.2', LsnM Transparency, LsMC 0x6 < ANC Transparency >, BTv 3.0, MicM Auto, Plcm M Enabled, Prim Unknown, Invl 11.25ms, srMd Disabled, spAM ContentDriven, HaSp No, HtSp No, CF 0x0 < >, initial
That doesn`t happen in the 27. Maybe a fallback event?
If I had to speculate I would say that the flag 0x40 makes the fallback not be triggered. And that was what makes versions 1.x.x keep working.
The versions 3.x.x probably are already working in the initial hid handshake, not needing the fallback action.
Well... I would love to have access to the code...