[FEN "2Q1Q3/2Q4Q/Q4Q2/3k4/Q5Q1/1R6/B1NBQ3/K2R1N2 w - - 0 1"]
How many mates?
What was black's last move?
hint: this construction (by Anthony Stewart Mackay Dickens, published in The Problemist, Jan 1970) ties the record (set by Nenad Petrovic, Sahovski Vjesnika, 1947), using fewer units.
Happy counting. But, frankly, I doubt this is useful to you...
This is certainly a remarkable diagram from a constructionist perspective, but it was never intended to improve a player's vision (in fact, if anything, such a visualization exercise will lead to blindness).
Stronger players may solve this faster (the guy in the video offers no statistical analysis), but solving faster will not put us among them.
Personally, I doubt that such unnatural mates in one can be a useful training exercise to improve your "vision."
I have no doubt that solving simple combinations and mates (e.g., Polgar's 5334 puzzles book) -- until the common patterns become automatic -- will improve your tactical ability.
A diet of endgames is even better!
Reading miniature games without a board might be the best ways to benefit your visualization (and your opening repertoire).
Another good way to improve visualization is to attempt "word" problems without a board.
For example, consider this problem from Sam Loyd (which I will paraphrase below)...
You took a new job, and just relocated into the wilderness. Your chessboard and computer will not arrive for a week or so. Despite that, you just received a certified letter from your former boss, with whom you reluctantly agreed to keep in contact by way of a correspondence chess game. This arrive early, no doubt, with the intent to put you at some disadvantage. But, no worries -- you have three days to answer the opening move, and who needs a chessboard to answer the opening move?
In the letter, your former boss boasts of needing to make it challenging.
"I play 1. f3," he double-exclaims... and there's more... he also sends the following list of conditional moves: "if legal, I play 2. Kf2, 3. Kg3, and 4. Kh4."
Technically, you need only answer the first move within three days, but despite having no chessboard, you have a strong incentive to spare your cheap opponent the cost of another postage stamp.
Can you find a series of moves which checkmates your opponent on your fourth move?
It is possible, but to find it, you'll need to unlearn everything you know about chess.
Just remember: finding the solution is not the point -- the whole point is to search for mating ideas of your own, and visualize the positions (to confirm your ideas without shuffling wood).