In practice, that doesn't help you much, because proving the existence of a watermark without revealing the methodology behind the insertion (say, for example, when you go into court to sue someone for piracy) is different problem than preventing an attacker from detecting the watermark on his own and manipulating it. This paper (and this paper and this paper, and this paper, and lots more, too) is an example of the level of sophistication detecting digital watermarks has reached. The only real long-term solution is tighter control of the formats that media are distributed in - watermarking is just a new battlefield for another endless arms race between publishers and pirates...
It's not simply a symmetrical arms race. In ordinary cryptography, the code-makers left the code-breakers in the dust many years ago, and the same thing is happening in steganography for the same underlying mathematical reasons.