Nope.
A PAO (or a PA, or a single-association system, or an LPAAO or whatever other structure you choose) where each element represents one card is a single-card system. One card’s information is represented by one intentionally translated mnemomic element. (The Person and the Action and the Object.) You have to accurately encode and recall 52 specific intentional mnemomic elements to memorize a deck of cards using a single card PAO. The PAO structure itself is just a way to organize those elements into scenes. But it’s not the content of the SCENES or how many cards they contain that determines the X in the “X-Digit System.”
If ToaD is the [Ace 
] and “CaRRying” is the [7 
] and a “NeT” is the [2 ] then the PAO scene “Toad carrying a net” gives me 3 cards worth of info in one scene that is composed by 3 intentional elements.
A 2-card system is where you compress two cards into a single intentionally created mnemonic element. This means you only need to use 26 specific intentional elements to memorize a deck. (The tradeoff is there are 2704 possible card pairs to create associations with.)
If “TRuCK” represents the card pair [Ace ][7 
] “SToNe” represents [Ace ][2 ] and “CaKe” represents [7 
][King ] then making a scene with these three elements gives me 6 cards info in one scene.
In a hypothetical 3-card system, you could combine 3 cards worth of information into a single element, similar to ToaD representing one card in the single-card PAO, or TRuCK representing 2 cards in the bigger system. This would be awesome because then you just need 17 elements to encode a deck. You could do this with like 6 scenes. But the problem is that there are well over 130,000 combinations of three cards. This is practically impossible to learn and train to fluency. Also practically impossible to create unique images for all those entries so as to not be confused between many of them.
The terminology may just seem like semantics but it can be important when communicating about how systems are constructed and used.
A system that uses a person-action-adjective-object-food-vehicle structure where each element represents one card is still just a “single card system.” It’s not a 6-card system. It is just structured into 6-element “PAAOFV” scenes.
This would likely be HARDER to construct and retain at speed than a simpler PAO or even a 2-element PA and it would require twice as many associations to learn and train to fluency. This is why I usually advocate for keeping scene structure simple. You don’t gain any benefit of data compression or reduction of elements to remember unless you can encode more cards across less ELEMENTS like with a 2-card system. If you don’t want to jump to the 2-card system with its thousand(s) or images to learn, then keep it simple since no matter how you structure it, you have to keep track of 52 elements.
Here’s more on that idea of clarity in terminology: