Let the Solution contain 2; then changing that to 4, 8 or 16 will not introduce new factors and hence we can replace 2 with 4, 8 or 16 get more new Solutions, which is against the Uniqueness Constraint; Hence Either all A = { 2, 4, 8, 16 } are Part of the Solution or all are not Part of the Solution.
Likewise for B = { 3, 9, 27 }
Likewise for C = { 5, 25 }
Likewise for D = { 6, 12, 24, 18 }
Consider E = { 1, 17, 19, 23 } which have no common factors with other numbers between 1 & 28. If one of these numbers is Part of the Solution, we can replace that with some other to get more Solutions; Hence Either all 1, 17, 19 & 23 are Part of the Solution or all are not Part of the Solution.
With E in the Solution, the other 2 numbers can have multiple Possibilities, which is against the Uniqueness Constraint : we can ignore E.
With A in the Solution, we can replace that with D, Example A+C is same as D+C : we can ignore A & D.
We are left with:
7 10 11 13 14 15 20 21 22 26 28
With 7&14 in the Solution, it is same as 7&28 : Ignore 7.
With only 11 or 13 and not 22 or 26, we can replace 11 or 13 with some other number from E : Ignore these.
We are left with Solution given by Athin earlier:
10 14 15 20 21 28
[[ Lot of hand-waving involved in my "Proof" !!!! ]]