Yes, the phys.org article includes a link to the original paper. (Did I say or imply otherwise? If so, I screwed up and I apologize. I didn't mean to.)
The GrrlScientist blog entry is better in (for instance) the following ways. It gives more context (e.g., Lorenz's earlier observation about formerly-captive animals running on wheels). It gives more information about the frequency of wheel-running by different kinds of animal. It is more accurate on at least one important point (article and GrrlScientist say 200k animal visits recorded of which 12k used the wheel; phys.org says 200k instances of animals using the wheel). It mentions, where phys.org doesn't, a very important finding: wheel-running continued even after the experimenters stopped putting food near the wheel, so it wasn't just that animals run on the wheel because they've found food there and think getting into the wheel will somehow provide more. It describes the article's comparison to wheel running in lab mice.
On the other hand, it's also more chattily verbose, which might be a good thing or a bad according to taste.
What's problematic about phys.org is that it's scarcely ever the best source for anything. Much of its content is recycled press releases, exactly the same as you can get from (usually) the website of the institution where the work was done but often with some illustrations removed and extra advertising added. Sometimes (as here) phys.org does actually write its own material, but (as here) it's usually more superficial and less informative than one can get from elsewhere.
There may be cases in which the phys.org writeup is (1) not just recycled from someone else and (2) the best available account of the matter for some plausible audience. If so, I just happen never to have encountered any.
[EDITED to add: I am not alone in thinking HN would be better linking to better sources than phys.org's blogspam. See, e.g., this lengthy collation: https://news.ycombinator.com/item?id=4676259. Full disclosure: some of the critical comments listed there are also mine.]
The GrrlScientist blog entry is better in (for instance) the following ways. It gives more context (e.g., Lorenz's earlier observation about formerly-captive animals running on wheels). It gives more information about the frequency of wheel-running by different kinds of animal. It is more accurate on at least one important point (article and GrrlScientist say 200k animal visits recorded of which 12k used the wheel; phys.org says 200k instances of animals using the wheel). It mentions, where phys.org doesn't, a very important finding: wheel-running continued even after the experimenters stopped putting food near the wheel, so it wasn't just that animals run on the wheel because they've found food there and think getting into the wheel will somehow provide more. It describes the article's comparison to wheel running in lab mice.
On the other hand, it's also more chattily verbose, which might be a good thing or a bad according to taste.
What's problematic about phys.org is that it's scarcely ever the best source for anything. Much of its content is recycled press releases, exactly the same as you can get from (usually) the website of the institution where the work was done but often with some illustrations removed and extra advertising added. Sometimes (as here) phys.org does actually write its own material, but (as here) it's usually more superficial and less informative than one can get from elsewhere.
There may be cases in which the phys.org writeup is (1) not just recycled from someone else and (2) the best available account of the matter for some plausible audience. If so, I just happen never to have encountered any.
[EDITED to add: I am not alone in thinking HN would be better linking to better sources than phys.org's blogspam. See, e.g., this lengthy collation: https://news.ycombinator.com/item?id=4676259. Full disclosure: some of the critical comments listed there are also mine.]