Most of .EXE processes are not monitorable at all, but there are some that are full-stack instrumentable, if they are built with .NET, as should be the case for all 3 you mention. Some details on those 3:
In my case, I have manually instrumented some .NET .exe executables, and with great success in some of them. You will probably have to create custom services to get a view of what is going on, especially Purepaths involving network/database requests. It is sometimes a very tricky subject, as you might need to have a code understanding of what is going on, but in some cases I have used it in conjuction with "Method Hotsposts" to be able to get to Purepaths without knowing about the code.
you mean a "do not monitor" (if EXE=...) in Custom process monitoring rules did not work out as you expect, that this exe is no longer monitored? I would expect it to be, because it there for to define process monitoring rules if you don’t want to monitor all your processes automatically, or if you need to define an exception for specific processes.
The objective is not to detect the processes once configured in the cluster. The rules are working perfectly fine but the unwanted processes will remain there without any purpose.
e.g. Nginx is not supported in Windows but you will find them in the technology and processes.
In a similar way, we want that once the process is marked do not monitor so on the next restart of the host/OneAgent that process should not be detected by OneAgent.
The same should apply to the hosts.