The scientific method was devised so that the truth could be uncovered in a systematic way, using general principles that reduce the chances that you are making mistakes. That's why hypotheses must be falsifiable, that's why methodologies and experiments must be published, not just conclusions; one group of scientists should be able to look at the work from another group and replicate it to show that the work is valid and no errors were made.
This is what a group of climate scientists has attempted to do with the 2-3% of all peer-reviewed climate studies that are either neutral or that disagree with the scientific consensus that global warming is real and a human-caused phenomenon. This is what science is about, right? If an experiment can't be replicated or if errors are found, the conclusions become suspect and can possibly be considered invalidated (that depends on what the error is).
The result? Well, it either will be surprising or not depending on what your expectations of the quality of the science being done by 'climate contrarians' was, but let's just say that some common errors were discovered. The study was published in the journal of Theoretical and Applied Climatology, and the most common flaw that was found was:
Cherry picking was the most common characteristic [shared by the studies]. We found that many contrarian research papers omitted important contextual information or ignored key data that did not fit the research conclusions. [...]
We found that the ‘curve fitting’ approach also used in the Humlum paper is another common theme in contrarian climate research. ‘Curve fitting’ describes taking several different variables, usually with regular cycles, and stretching them out until the combination fits a given curve (in this case, temperature data). It’s a practice I discuss in my book, about which mathematician John von Neumann once said,
With four parameters I can fit an elephant, and with five I can make him wiggle his trunk. (source)
An example of this was found in a paper that claimed to be able to predict future climate, but the model only worked for a cherry picked period of 4,000 years. When applied to a larger period, it didn't work anymore. Why should we trust a model to forecast future climate if it can't even reproduce known past periods?
Another very telling fact about the 2-3% of papers that disagree with the consensus: They don't share a coherent, consistent alternative theory to human-caused climate change. Some papers will blame it on orbital cycles, others on oceanic cycles, some on the sun, etc. Yet the 97% of papers that agree on global warming are part of a cohesive theory that is overwhelmingly supported by scientific evidence. These papers don't all go in disparate directions...
That, by itself, is another red flag for the 'contrarians'.