Loved seeing [OutputType([PSObject])] in an entry this morning... that helps the help system document what your script produces. It's a shame it doesn't work well with custom type names (since those are a bit of a fake-out on the object), but it's an attention to detail I appreciate.
I am seeing a little bit of misunderstandings. Keep in mind that the lastLogonTimestamp attribute in AD is the one that replicates, although there is a long possible delay in that replication. There are other "last logged on" attributes that don't replicate so you can't rely on them unless you're querying every DC (pretty inefficient).
Hey, one thing to think about: sometimes simpler is better. For example, instead of adding a dozen lines to check and see if a module exists and can be loaded, just add a #requires comment for that module. Let the shell do that work and spew an error if the module isn't present. It'll even force-load the module into memory. Saves lots of steps.
Hey, don't declare functions as global:Do-This. It's a neat trick, but you're polluting the shell's global scope. Plan to write in-scope functions and make them a script module, so they can be loaded and unloaded. From the Games perspective, "whatever," but in the real world... don't pollute the global scope.
A comment I saw: "You should check to make sure the module isn't loaded before loading it again." Disagree. The shell does this for you when you use Import-Module. But, doc your module dependency in a #requires, and you won't have to worry about the module. In fact, the whole theme of "checking to see if the AD module is loaded" appears to be a major point of commenting. I'm a fan of "easier" and a 1-line #requires -module ActiveDirectory is far easier to write and maintain than, say, and entire function designed specifically to load the ActiveDirectory module.
