For example, from the OP:
> Last week, the Pennington Biomedical Research Center in Louisiana announced the findings of a large new study on exercise in children that determined that lack of physical activity “is the biggest predictor of childhood obesity around the world.”
> The news release contained a disclosure: “This research was funded by The Coca-Cola Company.”
This is the link to the press release, which contains the Coca Cola disclosure in the last line:
https://www.pbrc.edu/news/?ArticleID=284
The press release links to the PubMed page for the study: http://www.ncbi.nlm.nih.gov/pubmed/26173093
...Maybe I'm missing it but I don't see the Coca-Cola disclosure in the PubMed abstract. However, the Coca-Cola mention is in the full text of the study: http://onlinelibrary.wiley.com/enhanced/doi/10.1002/oby.2115...
(note: this above link is visible for me because I'm on a university network. It may not work for you otherwise)
On the other hand, the full text of the study also mentions the ClinicalTrials.gov ID number: NCT01722500.
Which means it can be found here: https://clinicaltrials.gov/ct2/show/NCT01722500
So maybe the data collection approach should begin with ClincalTrials.gov, which does have endpoints intended for convenient bulk data analysis: https://clinicaltrials.gov/ct2/resources/download