Zeev Hed, the whole Wave structure is based on the premise that valuable bits (of various kinds) can be moved sufficiently securely to make it worthwhile.
BUT
1. No one claims failure is inconceivable. Just that it will fail well if it does so (meaning the problem will be isolated to the errant machine not the system).
2. Where the probability of failure and its cost are sufficiently low, insurance is capable of addressing the risk of loss.
3. I am not clear that the TPMs contain a clock. EMBASSY did, but currently the open channel resides with the TPM substructure.
4. There is little more than circumstantial data to support Wave's view of the near watertightness of its system. EDS tested it a year or two back. There have been one or two trials in Europe and Puerto Rico. But these were with the EMBASSY chip. Now IBM, Intel, NSM and Infineon are prepared to associate themselves with Wave's claims (some more loosely than others) in the TPM space. But there's little science to work with.