[danbruc]

Even if it were accurate enough, I think plate tectonics moves the crust around enough such that one would have a hard time finding a piece of rock that has been more or less always in the same depth. And I have no idea how static the core is, maybe even it undergoes some mixing process over long enough periods of time.

[dredmorbius]

If there are any such stable deep-crust locations, they're most likely located under a continental location rather than oceanic, if my understanding is correct. Ocean basins reform at a higher rate (and hence: are more-recently-exposed crust) than continental plates.

The very oldest regions are two locations, presently in South Africa and Australia, previously joined, dating back several billions of years.

The Jack Hills region in Australia has been dated to 4.4 billions of years (the Earth itself is 4.5 billion):

https://www.nationalgeographic.com/news/2014/2/140224-oldest...

A region presently under Greenland was dated to 3.8 billion years, in 2007:

https://www.newscientist.com/article/dn11438-oldest-chunk-of...

The Nuvvuagittuq Greenstone Belt, near Hudson Bay, Canada, has been dated to between 3.7 and 4.3 billion years old:

https://en.wikipedia.org/wiki/Nuvvuagittuq_Greenstone_Belt

All of these are surface rock. My thought is that drilling from these locations might find very old subsurface structures as well. Though whether any relativistic time dilation could be observed is hard to say.