First I need to point out that the above mentioned file is *not* a study. It's a preliminary report.
PhotoReading has been around since 1986 and yes initially we did a lot of research. Research that cannot be presented because it's going to be bias. The people who participated, wanted to participate. There is no way one can be neutral. Have you actually read the report? The primary subject of the report, trainee & researcher are the same person. That is also considered a bias.
The problem with research you cannot take the individual out of the research. Just like students do well in some subjects at school and fail dismally at others. Their enthusiasm and interest will determine their grade.
And when it comes to reading alone, 49% of the population in each of the English speaking countries that I've researched, cannot read to a 9th grade level. How can any research be accurate unless we manage to source people with same educational levels (since the material to be tested would require an unbiased interest), same measure of interest in the subject they are being tested. After all someone with an interest in cooking and zero interest in mechanics would do remarkably better on any test about cooking than mechanics. So the question in research then is were the subjects biased with an interest in the topic they were tested on (if the test has a positive outcome) or disinterest if negative.
I think when you consider the diversity of scores in an exam where among students had the same teacher, same demographics, same age group, even if you managed to exclude those who can't make the grade reading on par to the rest of the age group, you'll find individual bias towards the subject will always be a problem with validating the research.