Unfortunately we don't know how, if ever, the Smith family reacted to those e-fits. They were 4 adults and 2 teens. Could it be 3 thought Mr Slim plausible and 3 Mr Flat ?
It's a mystery all right.
This 'evolutionary system' (silly name, isn't it?) is so-called after part of the process in the e-fits' composition where the witness is shown a very large numbers of pictures of faces, and narrows the collection down to those which bear closest resemblance to the suspect as they remember him (or her).
Then the final picture is composed by the e-fit 'artist' out of those chosen images.
The result is a more real looking picture which is understood to have a much more powerful impact on the viewer because it contains elements of the 'internal features of the face' (whatever they are exactly), in contrast to the so-called photo fit police used to do which was essentially a jigsaw piece composed of eyes, nose, mouth etc. giving rise normally to a very odd looking facial image indeed (Used to give me nightmares after watching Crimewatch as a child!).
I wonder, if the so-called evolutionary technique was the one used with the Smith family, how they did it exactly. Would the individual family members all have been shown scores of pictures from which to select those most similar to their view of the suspect? Or would they have been shown images collectively?
Perhaps they worked separately and this in part would explain the disparity?
