Many companies are invited to participate, but choose not to. That's not the site's problem, nor the reviewers'. It might be helpful to add in the beginning of such reviews who was invited, and who specifically declined.
On a single-board sample, there is no way to test for general quality issues. Ahead of any review, on any site, for any product, readers should add "Assuming it works;" that is something we have to do. Unless a review indicates that the product possesses a liar label, there's really no way to meaningfully comment on QA issues.
As far as "questionable test results," did you expect the benchmarks to show large differences? In all the years I've been reading motherboard reviews, I don't recall seeing big differences, and often skip over the benchmarks because I don't believe they're going to show me anything remarkable.