I love that dog.
Anyway, I encourage you all to compile a coin using that insanely high value and see what happens every time you hit a difficulty retarget.
I imagine a value bnProofOfWorkLimit == 0 is not good.
But it will be == 0. It's actually pretty amusing to read the code for >> :
00416 CBigNum& operator>>=(unsigned int shift)
00417 {
00418 // Note: BN_rshift segfaults on 64-bit if 2^shift is greater than the number
00419 // if built on ubuntu 9.04 or 9.10, probably depends on version of OpenSSL
00420 CBigNum a = 1;
00421 a <<= shift;
00422 if (BN_cmp(&a, this) > 0)
00423 {
00424 *this = 0;
00425 return *this;
00426 }
00427
00428 if (!BN_rshift(this, this, shift))
00429 throw bignum_error("CBigNum:operator>>= : BN_rshift failed");
00430 return *this;
00431 }