And if this is such an important issue, and so many people remain to be convinced, wouldn't it behoove researchers to make their research as transparent, clear and reproducible as possible? Seems that can only help the debate move forward in a productive manner.

UPDATE: I followed your link, and things that stand out to me:

- Images showing station concentration shows only in fairly recent history that stations are global. Before then they're mostly concentrated in the US and Europe, and the station data only goes back to the mid 1800s, and is pretty sparse up until the last few decades. It is unclear how such sparse temperature data can give us a clear understanding of how world temperatures have changed over the past few centuries due to industrialization.

- The unfiltered 'all' category is not easily accessible. I'd either have to download a whopping 2.9G tar, or visit the directory that does not seem to load very quickly. I'm thinking it is just a flat dump of the files without hierarchical organization, so FTP tries to fetch everything.

- The accessible 'yearly' category readme.txt states the data is not raw, but 'merged together' and subjected to 'quality assurance review' just like the monthly counterpart, which I assume is the 'all' category. So, even if I did download the 2.9G tar, I cannot get the raw readings from the ground stations.

This is the sort of lack of clarity and transparency that does not improve the skeptic's perception on the matter.