Most people shooting landscapes do not do it to make them memorable, they shoot it because they are compelled to express what they feel about the world and humanity, and landscapes are their chosen medium. Just like mountain climbers climb mountains "because they are there". So whether landscapes are memorable or not carries very little weight in my case, even assuming the study is right.
Now, the study. It reminds me of those countless studies with "revolutionary" discoveries of the type "if you eat too much, your chance of getting obese doubles or triples". Another aspect of this study, to use analogy from the computing world, would be GIGO (garbage in, garbage out). That is, study conclusions are heavily influenced by the assumptions made and tools used in designing it.
For instance, if you use a generic postcard photo, as the one used in the article, of a mountain against blue sky, it is been seen by everyone millions of times before and is as "memorable" as if you turned the lights off or locked the door leaving the house. Compare that with human pictures in the study: each one represents a very unique or unexpected situation. So, generic vs. unique... hmmm... which one will I remember better? I see a Nobel prize in the making here for answering such a deep question.
Even the authors admit it: "...natural landscapes... can be memorable if they feature an unexpected element..." No shit, Sherlock!