Summing wildcard matches while subtracting wildcard matches of the previous rows

Hello and thanks in advance for the help!

I have a column of cells each consisting of a list of substrings separated by a comma. next to this column is another column with numerical values. Example:

A | B
a,b | 3
a,d | 15
c,d,b | 2
a,b,c | 8
c,a | 6
d | 9
b | 11
a | 4
a | 2
c,a,b | 1
a,d | 7
b,c | 4
c,d | 6


The substrings in col A are ranked:
1 a
2 b
3 c
4 d

Now, i need to sum column B for each of the 4 substrings, but making sure not to count the cells where it appeared alongside a higher ranked substring. I already have a formula that counts the instances of each substring when not appearing with a higher ranked substring (column E in example file) using this formula: {=SUM(N(MMULT(1-ISERR(FIND(TRANSPOSE(D$1:D2),A$1:A$13)),2^ROW(D$1:D2))=2^ROWS(D$1:D2)))}. In column F of the example file, I show what the results would look like for what I'm looking for.

Would appreciate some help on how to solve this.

Example spreadsheet attached

Thanks