Sune et al/on soapboxMy present understanding and opinion is the satellite data is the new "open" data source for air temperature. I have not read any accounts of Dr. Spencer hiding his algorithms or data.It has coverage that you can't possibly get with a ground based network. Ground based stations with the long term accuracy, stability, and durability are very expensive - and will always be subjected to changes in the local environment. You are also dealing with thousands of ground based sensors to monitor and maintain versus a satellite (or a dozen) with one sensor to maintain which is used for land and ocean temps. Whatever error band the satellite has is consistent for all measurements - unlike the ground/sea based network.
There are - as extremely well documented on this site - issues with proxy based reconstructions of past temperature as well as the present ground/ocean network. While it is "fun" to consider what climate was like in the past thousand years, it appears to me that the reconstructions do not have the accuracy or resolution required for more than general statements. Trying to extract more precision out of them with statistical algorithms is something I do not believe is possible. If "smart" folks with Phd's can only get "reasonably" looking global temperature reconstructions by picking out a particular tree ring on a mountain range in the western USA and giving it several hundred times the weight of the other proxies in the reconstruction (and that is just plain silly IMHO), then I don't think it can be done. While Steve has never said it is not possible, he has pointed out the flaws in the reconstructions used by IPCC.
The present ground network was not designed for the the purpose of monitoring climate change. The stations do well what they were designed to do - monitor local weather conditions for (mostly) agricultural purposes - not climate. Again - trying to extract a small climate signal out of them by creating models to "correct" the readings is not consistent or reasonable with their purpose and design. I think it is wishful "thinking" to claim that a small climate signal can be extracted with precision. Too many variables - too little measurement precision. Again, Steve has never claimed it cannot be done, but he also has never claimed that it is possible, either.My opinion./off soapbox -->