The key takeaway:

> Noting that “buoy data have been proven to be more accurate than ship data,” the new study applies a new “bias correction” to address the difference between them.

..by adding a correction to the more accurate buoy data, generating a combined dataset that does not match trends in the satellite data, gold-standard terrestrial data, balloon data, or oceanic temperature data.

It also reduces the long-term temperature trend to the pre-1950 trend (i.e., pre-anthropogenic trend).

This is a paper that requires a significant degree of corroboration rather than simple endorsement because of its publication.

Pre-1950 trend? I don't get it. What happened in 1950?

1950 is generally accepted by both AGW skeptics and proponents as a time before which human influence on climate was not discernable. That is, the warming trend before 1950 was natural, not human-caused.

AGW skeptics assert this places a burden of proof on AGW proponents to show that warming since 1950 is beyond what could naturally occur.