Conditional measure to avoid sum duplicates

[mojo_flex]

Hi,
I would like some help creating a measure that avoids summing duplicates and incorporates some conditional logic.

I'm very new to DAX and so it may be that what I am attempting is impossible or better achieved in an alternate way. Advice, greatly appreciated!

Given something similar to the following table:

I'd like to create a measure that avoids duplicates across all columns for CAT = A but only avoids duplicates across CAT and VALUE for CAT = B or C.
Essentially the measure should sum the following:

The following code successfully avoids duplicates across all columns but I am unsure how to ammend it to achieve the conditional logic described.

[measure] = 
SUMX(
    SUMMARIZE(
        T,
        T[DATE],
        T[CAT]
        T[VALUE]
    ),
    T[VALUE]
)

The above essentially achieves:

However for CAT B and CAT C I guess the following should be applied:

[measure] = 
SUMX(
    SUMMARIZE(
        T,
        T[CAT],
        T[VALUE]
    ),
    T[VALUE]
)

Many thanks in advance!

[ERD]

@mojo_flex , can you, please, show the result you want to achieve?

[mojo_flex]

@ERD further explanation added...

[ERD · Accepted solution]

@mojo_flex , one of the ways to achieve this is to use the virtual tables you've created:

conditional_sum =
VAR t_A = SUMMARIZE ( T, T[DATE], T[CAT], T[Value] )
VAR t_BC = SUMMARIZE ( T, T[CAT], T[VALUE] )
VAR current_cat = SELECTEDVALUE ( T[Cat] )
RETURN
    IF ( current_cat = "A", SUMX ( t_A, [Value] ), SUMX ( t_BC, [Value] ) )

Is this what you want to achieve?

[mojo_flex]

Answers my question perfectly. Thank you for taking the time to help me.
I was wondering if splitting my table was going to be neccessary as I wasn't aware of the virtual table function. I really appreciate the pointers!