Think I am on the way to solving the problem. There is a difference apparently, which I understand (in maybe a slightly naive manner) to be that Kodachrome transmits more infrared and therefore there is the risk of greater contrast, biased towards the shadows (effect less noticeable with highlights). This apparently especially affects Digital ICE scratch and dust reducing plugin. Presumably this risk is such that some software, veering towards the professional end of the market, has to take this into account.
However, I use Digital ICE for all my scans and on normal setting it seems to have no real effect on contrast, plus it saves a lot of cleaning up of the scanned image (on fine there is a lot of blurring, so guess fine is for badly 'dirty' images and a lot of work with the unsharp mask). Confirming this finding is this article
Scanning 35 mm Slides. There is no real difference when using Digital ICE, and the comparison with Ektachrome is interesting.
Bottom line is that really I should relax and judge my results by what I see on screen, not in theory. Where the theory helps is to analyse and correct the odd rogue scan.