@AC w/r/t "Fritz Chip"
A name I hadn't heard, but which turns out refer to the Trusted Platform Module (TPM). It seems to me that this makes some of your assertions questionable (IMHO)
Quick whistlestop, the TPM provides three (count them) sets of functionality : Public key crypto functions (RSA, SHA-1, HMAC) , trusted boot, and initialisation and management.
Since the user (and there are GPL'd linux drivers available for the module) may define exactly what configuration is to be 'trusted', there is nothing on the 'Fritz' * chip to prevent me from booting linux, starting a VM, attaching a debugger and using this environment to host another OS.
Additionally, I can provide my VM with a virtual TPM (e.g. one in software) http://www.usenix.org/events/sec06/tech/full_papers/berger/berger_html/vtpm06.html
The utility of this is debatable, but the point is that I can set up an environment with a TPM enabled OS running in an environment that I control completely, including the internals of the TPM. (Whats that ? Newer OSs detect when they're virtualised ? That's an old, old arms race.)
There is NO way for the TPM to reject an OS, a system configuration, or a piece of hardware because it isn't certified by a third party, none, zip, zilch, nada, bugger all.
From the horses mouth :
"Can the Trusted Platform Module control what software runs?
No. There is no ability to do this."
"Does TCG require that software be certified to run on a TCG-enabled platform?
The TCG design does not have any requirement that software be “certified” in order to use it."
Really, it just doesn't work like that. (Cue hysterical freetards shouting that TPM, and trusted computing in general are evil and that the sky is falling, t'aint so, increase Ritalin and drink less coffee)
Is it possible that you are getting the TPM mixed up with Microsoft's NGSCB (Next Generation Secured Computing Base), formerly known as Palladium, and which relies on a superset of the TPM hardware functionality, including things like Memory Curtaining, and which is indeed far more genuinely sinister ? Or maybe Intel TXT ? Or just possibly something else entirely ?
* Apparently, after US Senator Ernest Frederick "Fritz" Hollings, a sock puppet for the RIAA/MPAA and their pigopolist friends who lobbied, unsuccessfully, for a number of draconian anti copying measures which, had they been passed, would have made US consumer electronics manufacturers even less competitive with their Chinese, Malaysian, &c brethren than they are already. One of which was to include the TPM chip in all media devices. Presumably either he, or indeed the MPAA/RIAA had either misinterpreted the functions of the TPM, or had bought into MSFTs much more Machiavellian Palladium vision.