Microsoft say that the main features of TPMs are that they [1]:
> Generate, store, and limit the use of cryptographic keys.
> Use TPM technology for platform device authentication by using the TPM’s unique RSA key, which is burned into itself.
> Help ensure platform integrity by taking and storing security measurements.
One can, in principle, imagine situations á la Apple's T2 chip whereby this could be very useful to the end user -- for example, in hardware rate-limiting whole drive encryption decryption requests. Microsoft don't actually state this as a potential use-case. They go for the rather more prosaic
> Antimalware software can use the boot measurements of the operating system start state to prove the integrity of a computer running Windows 10 or Windows Server 2016. These measurements include the launch of Hyper-V to test that datacenters using virtualization are not running untrusted hypervisors. With BitLocker Network Unlock, IT administrators can push an update without concerns that a computer is waiting for PIN entry.
The rest of the page then goes on about hardware attestation. In reality, I am increasingly convinced that this is all an elaborate DRM scheme, similar to what they integrated in the XBox, with its on-chip crypto. I think we will see increasingly user-hostile, but more "transparent" DRM schemes based around this idea, and continue the cat-and-mouse game of "you are running this code in a VM and that is unauthorised for $MONEY_REASONS".
I'll stick to Linux, thanks.
[1] https://docs.microsoft.com/en-us/windows/security/informatio...