Thinkbox

The findings are a product of "meta-analysis", meaning they include no experimental testing on their own and include all of the possible trappings of the format (errors in original reports, publication biases, incompatibility of includes studies, etc.)
The report itself describes the method of analysis it's referencing, market mix modeling or "MMM", as "looking to explain the variation in an outcome we care about by correlating this variation with consistent changes in the inputs", but then goes on to suggest causality, saying "Advertising is a profitable driver of business growth"

"It should be noted however that much of this evidence comes from the realm of digital advertising. Much of the focus has been on improving viewability metrics and, as described above, the improvements can be seen and measured. Results such as these have led to recommendations about advertising and media practices. It should be noted though that this is a specific exposure situation and that the findings may not automatically generalise to all media."
"Peter Pynta from EEG based company Neuro-Insight suggested that visual attention only accounted for 15% of memory encoding and other factors such as storytelling emotion, frequency and context are commonly implicated in the remaining memory encoding. Other factors such as salience, distinctiveness and processing level are also likely to be involved."