Almost all mice & keyboards are 1000Hz USB 2.0 devices rather than 8000Hz USB 3.0. So a direct m/kb connection to PS5 is at best 1ms. However, the PS5 polling interval for controllers over USB is 4ms, so that's likely to be the same for m/kb.
Beloader has sub-1ms roundtrip pings over ethernet (home network, not internet). A XIM + Beloader connection is at worst 2ms with a XIM APEX Response Rate of 1000Hz. If anything, Beloader with XIM is twice as fast as native m/kb via USB.
If you were using Beloader over Wifi, your XIM (250Hz Response Rate to avoid desync) + Beloader (limited to 250Hz) connection would be about 5-8ms.
If you were using PS Remote Play app on PC or Mac with XIM, your connection would be roughly 14 to 51ms. That might seem slow, but still feels pretty smooth and responsive in most games.
Keep in mind that the PS5 games run at 120FPS at best, or 8.33ms. So as long as input from m/kb or Beloader lands within that frame (twice with native m/kb, four times with Beloader), you'll get maximum input speed.
Consoles are not designed for 1000Hz input, they're designed for 125 to 250Hz input. Any faster than that and you introduce problems like mouse jitter, audio garbling, dropped frames and aim assist issues. So if you set your APEX to run at 250Hz, it'll perform roughly on-par with native m/kb, but with strong aim assist.
Thank you very much for the very good explanation. My friend claims that he has a larger ping/latency in COD with beloader (via Ethernet hub). Without a beloader he has supposedly a low ping/latency in the game. Can that be true? I can't imagine it, because the loader works in the home network.