NP, I'll try and work more on it this weekend and will release the code to you, and then a cleaned up copy of the code for everyone else here on the forum that does NOT divulge any of your insider secrets, etc. That way people can see the basics of how I did it, and none of your personal details are exposed.
I would define rude scrapes if it's the same request or similar in a short amount of time. For example, I run my own phone line at my house on Asterisk and my Linux box, and when I call back to friends in Austin, TX it's [always?] the same rate for anywhere in Austin that I call. Thus if I were to query the ratecenter that I use (sooner or later I'll be adding VoicePulse onto my line-up, and breaking out their Macro script as well), I would take the easy way out and only query area code 512. If it turns out that another area that uses 512 (Bastrop for example?) is not equal to Austin, then I could target the area code and the prefix of the # if I had to.
In summary, caching is our friend; and (in theory) doesn't trouble the administrators that are sifting through log files and noticing that 1 IP is making a lot of similar requests in a short amount of time. :)