While i do support terracoin i think the issue of someone poining asics at the network and thrus forcing up the difficulty of a reletively new coin shows a major flaw in most alt-coin designs - using the same algorythm as bitcoin, imho any alt coin that wants to stand the test of time and actually be taken seriously is going to need a difference algo - take litecoin as the perfect example, different algo and it's followed the same mining evolution as bitcoin.
I agree. History has now shown that starting an alt coin that uses the same hashing algorithm as an existing, strong, coin is a bad choice. We now have had several examples of this.
The first was Namecoin, where the difficulty went way up then mining power left and it was stuck for several months at super-high difficulty.
The next was CoildCoin (CLC). It had merged-mining right from the start in an attempt to have a higher hashrate. However, Luke-Jr (who controled the Elgius pool at the time) set the pool to merge-mine it and not allow anyone elses blocks or allow any transactions in his blocks. Since Elgius had a much higher hashrate then the rest of the network, he was successful and the network stalled.
For Terracoin (TRC) a single person with a bitcoin ASIC starts mining and causes the difficulty to fluctuate wildly (first difficulty calculation method) or causes it to go up and stay up (second difficulty calculation method). That person had over 50% of the hashpower and could have done much more damage (and still can if they want to). They could mine their blocks exclusively, ignoreing any blocks found by anyone else, thus getting all the block rewards, excluding others transactions, and doing double spends at will.
re-using the same hashing algorithm sounds like a good idea, it takes less effort, you don't have to worry about introducing new bugs, and you can take advantage of all the optimized software/hardware to get your hashrate up. However, an attacker has access to the same software/hardware that you do, may already have a lot of it, and, often has a vested interest in a different coin succeeding so have a reason to use those resources to attack a new chain.
This is certainly true for sha256 and is mostly likely true for scrypt now (if not, certainly when scrypt fpga's come out it will by).