That is not true - for example, the Redwing Zuni / Bassoon test in 1956 yielded 3.5Mt, of which 85% was from fusion, due to use of lead for the tamper around th

That is not true - for example, the Redwing Zuni / Bassoon test in 1956 yielded 3.5Mt, of which 85% was from fusion, due to use of lead for the tamper around the third stage. On the other hand, if primarily looking to maximize yield for cost, one can use U-238 instead - the higher energy neutrons from the D-T fusion will cause fission (although not a chain reaction) in the tamper as well. Redwing Tewa / Bassoon Prime test achieved 5Mt with this change in the same the test series and the design was refined ultimately to the B-41 gravity bomb with 25Mt yield. We know how to make devices with primarily fusion yield, it's more that they're not as effective as weapons.