only if you're ok with never buying any physical goods with bitcoin. the weak point is the exchange between bitcoins and classical currencies or physical goods.
Running through TOR doesn't actually affect this. After the first node, the only data on who is giving money to who is the bitcoin public keys. The only way to attach those keys to a person is to have access to a node that is 1 degree removed, in which case you have the IP adress from TCP, or look through the. block chain to see who else the bitcoin ID you are looking into traded with. The first aproach is not common, and TOR will not help on the second.
When you are a business fearing regulation, this is all irrelevent, because you need to make public what customers get in return for sending you bitcoins, as well as your public key, so customers know who to pay.
It would be interesting to see a truly anonamaus bitcoin system. You would need invalidate a coin for which you do know a secret, and create a new coin for which someone else knows a secret, in such a way that when the other person proves they know the secret no one can figure out that the coin they know about is the one you gave them. Giving all of the things already possible, this seems like it is easy in theory. Doing it in a computationally efficient way is a different story.
When you are a business fearing regulation, this is all irrelevent, because you need to make public what customers get in return for sending you bitcoins, as well as your public key, so customers know who to pay.
It would be interesting to see a truly anonamaus bitcoin system. You would need invalidate a coin for which you do know a secret, and create a new coin for which someone else knows a secret, in such a way that when the other person proves they know the secret no one can figure out that the coin they know about is the one you gave them. Giving all of the things already possible, this seems like it is easy in theory. Doing it in a computationally efficient way is a different story.