Bitcoin is based on a solid algorithm, but is driven by a community that seems to seriously lack objectivness, probably because of vested interest into ever increasing exchange rates. I am not sure if they are not going to crash the whole thing out of blind and zealous faith
If I were you I would wait and see if the whole Bitcoin ecosystem does not collapse entirely in the coming days.
Maybe you can start by building one rig, and convince yourself that the profit is worth the trouble.
You should also make the experience of one or two difficulty increases, and factor the outcome into you profitability model.
In bitcoin denomination, you profit is going to decrease exponentially over time, so you will have to rely on a permanent and proportionnal (therefore exponential) increase of the BTC/USD exchange rate if you want to have a stable income in your native currency.
Now if you have alternate business models for your hardware, like creating a GPGPU cloud service or if you would like to help cancer research by running folding@home, then it's a different story.