Why only Open Access in Scopus and not all articles that are open access?

Because Elsevier are the sub-contractor for this.

It is good to observe important open science tools, such as DOAJ, ROAD, OpenAire, etc. being proposed here; but I worry that there is insufficient emphasis on Green OA. For example, only OpenAIRE will surface data about Green OA, yet OpenAIRE's coverage may be limited. Is probably necessary to seek inclusion of IRs or aggregations outside OpenAIRE infrastructure.

Publication sources cannot include Scopus. If Elsevier, subcontracted in this monitoring action, had its own open science agenda, they could alter their own criteria for inclusion of Open Access publications in Scopus and steer conclusions about OA. There is clear, immediate conflict of interest.

What is the CWTS method for Open Access identification? According to the note, this is a combination of 5 different sources (DOAJ, PMC, ROAD, CrossRef, and OpenAIRE). As mentioned above, these data can be obtained from a single services, such as Web of Science and Unpaywall. AGain, it is unclear why Clarivate's services has been replaced with the more complex Elsevier one. Will these data, matching algorithms, and other methods be made available?

I don't now what the CWTS method is, but I agree with the comment from Jon Tennant above (https://www.makingspeechestalk... ): Its troublesome that tax-money is used to reinvent the wheel when the data is readily available by other services.

Also, I hope that CWTS is better than the one currently applied by Elsevier/Scopus, as many OA articles are not reported OA in Scopus or not listed at all (examples can be provided). There is a great risk to greatly underestimate the role of OA by the currently proposed methodology.

Further to my comment above, I want to draw attention to a recent publication, in which two of the authors are from the CWTS: https://osf.io/preprints/socar... Here, this study uses the same sources, with the exception that it uses Google Scholar (free) instead of Scopus (paid). Why is this not considered here, when members of the same research group are using the method, and clearly to the same effect? They are able to adequately and accurately assess publisher-based OA proportions for the different 'types', and delineate the data also based on discipline and country.

Using this indicator will result in a very strong STEM-journal-English speaking bias. At least DOAB (Directory of Open Access Books) should be used in order to include Open Access books.

Monographs are absent or at least poorly represented, thus leaving SSH production clearly unrepresented. Strongly advise to use at least DOAB references.