Common Values in 3 Columns

You could use a number of SCAN or REDUCE type of LAMBDA functions to loop through each list but here is what I think is a little more general option.  I also assumed that each list could have duplicates themselves:

=LET(in_1, O1:O15, in_2, P1:P15, in_3, Q1:Q15,
     ustack, VSTACK(UNIQUE(in_1),UNIQUE(in_2),UNIQUE(in_3)),
     Duplicates, LAMBDA(array,[number],
                 LET(num, IF(ISOMITTED(number),1,number),
                     counts, MMULT(--(array=TRANSPOSE(array)),SEQUENCE(ROWS(array),,1,0)),
                     UNIQUE(FILTER(array,(array<>"")*(counts>=num),"no dups"))
                 )),
     Duplicates(ustack,3)
)

and you just need to have row 1 and row 2 match the actual input columns (both in range reference and number of input columns)  and then on row 8 that number 3 matches the number of columns.

Here it is in action: