I'm an amateur so I go by "rules" rather than "why". Said differently, I've been taught "what to think" rather than "how to think".
So I've been taught to avoid cascading switches. That said I have and it works. Let's play for a moment;
Availability: Let's
pretend any one device will go down once a year such that the availability of any one device is 99.7%. Add five (5) more devices inline and that's six (6) outages a year with an availability of what, 98.6%?
Repair: We've added five things (as well as additional cables, AC power adapters, outlets, circuit breakers, etc.) that could go wrong. Which
one is the culprit?
Latency: Let's pretend Ethernet latency to the router is 0.5 ms. Now let's stick five more devices (switches and cables) in between. That "could" go up to a couple of milliseconds (and it might not because I don't know how to calculate the benefit of "cut through" switching : -)
Now that would be just a "nit" in the world of 30ms Intetnet latencies but could be of interest if you're running an old fashioned client/server database application locally.
Congestion: On lower end switches could/do the gigabit ports connecting the switches inline become a
"pinch" point?
Impact: If all the switches were "home runned" a switch failure would knockout 6 cameras. If they were all cascaded / connected in serial a single switch failure would knock out anywhere from 6 to all 24 cameras.
Myself, I'm not seeing a significant issue. Bottom line you've gotta do what you gotta do. Worse case you can always go back and "home run" a switch or two as needed. Have you considered "home running" (not the cameras) but just (some of) the switches?