STARK proofs rely on crypto assumptions that are even older than the crypto assumptions bitcoin already relies on (collision-resistant hash vs elliptic curve DLP; see [here](https://user-images.githubusercontent.com/9424721/186959627-1f09a5d3-a4d5-4d06-980b-bd0791bd1543.png))

Regarding implementations, there are validity proof implementations that are about the same age as the Schnorr implementation that is used in bitcoin.




lightcoin

bitcoin

Can Validity Rollups Help Scale Bitcoin?

Sources - idk I work in IT/mathematics and have spend a lot of time with zk-proofs such as signatures and hash preimages (their derivatives specifically). I do think they look cryptographically sound for now - but the spirit of Bitcoin has been betting on standard and well established cryptography for a long time.  Modern zk-proofs do not have this long history of being unbroken yet 