PowerShell works on fundamentally different principles than do traditional Unix shells. Unix shells are for loosely coupling small "do one thing and do it well" programs. PowerShell can kinda do this, but its strengths lie in the "object streaming" features which aren't based on external processes at all but on loading .NET libraries with a special API, called cmdlets. In this respect PowerShell is a bit more like the VMS command processor, or using Python or Tcl as a shell. To add a command to a Unix shell, you need but drop an executable in your path. To add a command to PowerShell that takes advantage of PowerShell's features, you need to write a cmdlet in a language like C# and then install the cmdlet. This is also why PowerShell takes so long to boot up: it's probing for and loading installed cmdlets.

For these reasons I do not consider PowerShell to be directly comparable to Unix shells, but it is an essential part of the Windows toolkit because no one wants to use the MS-DOS-like cmd.exe for anything serious.