I actually just got a Lattice Semi board, which is not the most ideal FPGA for SHA256, but it certainly does fit in the price point I have, and maintains a lot of flexibility for the FPGA deployment I want to do (more generalized, but with the free power I can get with FPGA deployments on private energy producers, it's important to be able to recontribute back some stuff). I'd rather have one that I can re-purpose for radio, PCI and ethernet easily for $99. I also was really happy to find an American made board delivered domestically, and in really legit packaging.

You may also want to consider looking at optimizing http://cryptography.gmu.edu/athena/index.php?id=source_codes the SHA3 candidate cores released by the NSA, which might help you hedge towards a new future chain and (depending on where you live), maintain a domestic production economy for gear in the event of processor fragmentation into multiple chains across the planet. We'd do better taking FPGAs like this and putting them towards cracking Nagravision 2 and 3 (bottom line wise, anyhow).

I'm aiming on optimizing the newer SHA3 potentials (my current speculative guess for usefulness is http://keccak.noekeon.org/, because with this much sha256 hashing power out there, alternative chains are going to emerge.

FPGA is cool for BTC, but everyone is kinda rusty. And the real way to finish the tail end on the BTC is getting a few thousand 5W chips on top of large solar and wind arrays. It's just gotta be able the bake in sun and heat.