So I'm skeptical of estimates more than a millennium back in time until we have multiple proxies we can trust at that timescale, and can verify somehow the DC component of the estimates, or at least replicate them.

WSJ: Climate Emails: Science and Candor