Imagine a game in which you choose many cards with different A,B,C values.
Such as :
Card 1
A - 4
B - 5
C - 6
Card 2
A - 2
B - 7
C - 4
...
and so on..
To win the game you have to reach (or exceed) any of the two A,B,C attributes to a pre-defined value. For the sake of our example lets assume :
A - 100
B - 200
C - 50
Consider the straightforward algorithm when trading cards in which, each card has a value based on their contribution to the goal number (c attribute is 4 times more valueable than B attribute ) :
P ( Total value of a card ) = A / 100 + B / 200 + C / 50
Now suppose I introduce two new attributes D and E to each card, and a goal G on top of the ones already introduced.
Such as :
Card 1
A - 4
B - 5
C - 6
D - 3
E - 8
Card 2
A - 2
B - 7
C - 4
D - 12
E - 16
...
With the winning condition becoming based on :
C1 - A > 100
C2 - B > 200
C3 - C > 50
C4 - D / E > 0.5
How should one integrate the attributes D and E to the equation when calculating P ?
Based on the function $P$ given in your description, it seems that you want $P$ to map a 5-tuple $(A,B,C,D,E)$ to a number that roughly corresponds to the value of the card. If that's the case, the most obvious extension (in my mind) of the function $P$ that you give is to have $$P(A,B,C,D,E)=A / 100 + B / 200 + C / 50 + 2 D/E.$$ This definition of $P$ agrees with the one that you gave in the description in the sense that if C1, C2, C3, or C4 is satisfied, then $P(A,B,C,D,E)>1.$