This "IPC hobby" is often like a black hole: the more the merrier, but in case of tight budget, some shortcuts might be taken, however do take note of these considerations as you (he) might want to upgrade the system later on. For example, I strive to "cover" each camera with another camera, for different reasoning: you can always see whether that cam works (eg turn IR on/off remotely), or even better: see if someone is damaging it. In your case, you might end up with double amount of cams. On the other hand, in the current diagram there are "blind spots" where people can come (eg from garden side) and run straight to the house without being noticed. If there are windows to these walls, your plan isn't waterproof.
Which brings me to a more important question:
what is the objective of this exercise? Is it to "scare off" potential burglars, have "some" footage and feel a bit more "safe", or is this to have "24/7 intelligent video analysis" to see any activity? If it was to feel more safe, I would first insist to install a decent alarm system, with door/windows/area/PIR sensors before going into the cam level. They CAN complement each other, but cams do often proof to generate false alarms (eg cat passing by, shadows of all kinds etc).
We on this forum can formulate tons of advices, which devices, network diagrams etc - all with a certain price - but if you can share some background, it would help to "limit" the scope. You don't want Fort Knox advice right?
