Twitter has been touted for its myriad of potential uses — from social networking to news headline tracking to marketing yourself. To this growing list, you can now add a new item: remotely controlling your Mac. That’s what you can do with TweetMyMac, a Mac application available since September.

Imagine sending a Twitter message from wherever you happen to be, containing a command such as “open Safari,” and having the requested command carried out on your Mac back home. That’s the way TweetMyMac works.

To get this set up, you need two Twitter accounts: your main one and a special one set up just for TweetMyMac. Next, launch TweetMyMac and leave it running. Now, send any valid command via a direct message from your Twitter account to the TweetMyMac account. When TweetMyMac receives the message, the command is carried out.

There are currently 21 commands that are understood by TweetMyMac, ranging from Shutdown to Mute to Screensaver to Open [application]. There’s even a shell command that, assuming you know how to work with shell commands, opens the door to functions not included in TweetMyMac’s predefined list.

After playing with TweetMyMac over the past few weeks, I find myself vacillating among three different reactions: “Wow, this is truly amazing,” “Technologically impressive but, really, who needs it?” and “Gak, the implications are a bit scary.” Let me explain each reaction in some detail.

Wow, this is truly amazing. This was my first reaction. I was amazed that something like this was even possible — via Twitter. I began to consider possible practical uses.

Suppose I left for a trip and forgot to turn off my Mac. I could now turn it off via Twitter.

Or perhaps I left my Mac on deliberately, so that I could access it  for file sharing while on a trip. The only problem is that I forgot that I would need the Mac’s IP address. No problem, a TweetMyMac command sends me the needed address. Or perhaps I just want to quit my email program on my Mac, so that it doesn’t automatically download email from a POP account before I can grab the email from my iPhone.

I can confirm that all of these commands work as advertised. There’s usually a lag of about a minute or so, between when I send a command and when it takes effect. But it otherwise worked fine in my tests. As I said, amazing.

Technologically impressive but, really, who needs it? After getting over my initial enthusiastic response, reality slapped me in the face. My so-called practical examples sounded great, but how often would I really need to do these things? Almost never. And unless I left TweetMyMac running all the time, it might not even be running when I needed it.

On closer inspection, many of the commands seemed entirely worthless. I mean, why would I want to remotely launch an application on my Mac when I can’t further interact with the application in any way? I might want to do this if the application was required to remotely connect to my Mac in some other way. Otherwise, I can’t think of a reason. Actually, if I really wanted to control my Mac from my iPhone, I would likely be better off with a app such as Mocha VNC Lite.

Gak, the implications are a bit scary. For TweetMyMac to work, you have to set the TweetMyMac account to follow your regular account. It’s important that you do not  set TweetMyMac to follow other accounts. Why? Because any account that it follows can send commands to your Mac. This is where it starts to get a bit scary.

If you select to follow others with your TweetMyMac account, and they are aware of this, they could send commands to your Mac even while you are at home working. Imagine having Safari quit while you are in the middle of surfing — because someone remotely sent a quit command. It could happen.

“You’re exaggerating this risk,” I hear you saying, “because I have no intention of ever following anyone but myself.” But imagine this: You work in a semi-public place, such as a large office. You briefly get up from your desk to get a cup of coffee, leaving your Mac unattended and unlocked. Now anyone could come along and add their account to TweetMyMac’s “following” list. When you returned, you’d never detect this unless you actively checked.

At this point, the person with the newly added account can send commands to your Mac. Hopefully, someone would only do this for an “April Fools”-type practical joke — such as using the Say [phrase] command to have the Mac suddenly shout out an insult. Ha ha. But they could do something more serious.

The scariest scenario is the possibility that some hacker could figure out a way to gain this access even without you explicitly following them. I know this seems unlikely. But check Apple’s list of security updates; you’ll see that all sorts of unlikely events that are quite real.

Bottom line. TweetMyMac: a great new utility, a worthless piece of software, or a dangerous risk? Which is it? It’s a bit of all three. If it sounds intriguing, try it out and decide for yourself.