I am concerned that the 'Hammy conform' might be a problem. For example, there could be a situation where a base is 1mm away from conforming. If this is blocked then it would have to slide 39mm in the opposite direction. This might create some very odd results, particularly a vulnerable overlap overlap appearing.
Lawrence has stated the situation clearly. I am not sure you have clarified things very much Simon. Stating 'full base contact' is digging us in a bit deeper. I am now unsure if you would agree with the situations described below.
Two situations and how we are playing it:
1) One base on one base contact at impact, length of contact is 1mm. Conform is a slide of 39mm to ensure there is one base full contact. I.e. it is not an option to slide out of contact to be just an overlap.
2) Two bases of each side contact at impact, but are offset so that one of the bases only makes contact with 1mm of frontage. In this case, the slide to conform is 1mm. The base with 1mm contact has now become an overlap. The same number of bases fight, but there is one in full contact, one as an overlap.
Situation two is consistent with the angled contact where the base has hit the side of the enemy and conforms to overlap (as described in the rules)
(This has now now extended the argument. You started it by talking about full bases

)