SilverFast has the reason wrong.
The Wikipedia article on Kodachrome has it correct about the problems with scanning Kodachrome and the nature of developed Kodachrome:
"Many scanners use an additional infrared channel to detect defects, as the long wave infrared radiation passes through the film but not through dust particles. Dust, scratches, and fingerprints on the slide are typically detected and removed by a scanner's software. Kodachrome interacts with this infrared channel in two ways. The absorption of the cyan dye extends into the near infrared region, making this layer opaque to infrared radiation.[41] Kodachrome also has a pronounced relief image that can affect the infrared channel. These effects can sometimes cause a slight loss of sharpness in the scanned image when Digital ICE or a similar infrared channel dust removal function is used.[42]
Processing of Kodachrome films
Kodachrome, and other non-substantive films, unlike most color films, required complex processing that could not practicably be carried out by amateurs.[43] The process underwent four significant alterations since its inception.[44] The final version of the process, designated K-14, was introduced in 1974. The process was complex and exacting, requiring technicians with extensive chemistry training and large, complex machinery. This is because most color films contain dye couplers in the film itself; during development the couplers react with the developer to form the dyes that form the final image. Kodachrome film has no such couplers; instead the dyes are formed on the film by a complex processing sequence that required four different developers; one black and white developer, and three color developers. Normal color film requires just one developer. Also, processing Kodachrome film requires 8 or more tanks of processing chemicals, each of which must be precisely controlled for concentration, temperature and agitation, resulting in very complex processing equipment with precise chemical control, no small feat for small processing companies.[45]
The first step in the process was the removal of the antihalation backing with an alkaline solution and wash. The film was then developed using a developer containing phenidone and hydroquinone, which formed three superimposed negative images, one for each primary color.[44] After the first developer was washed out, the film underwent re-exposure and redevelopment. Re-exposure fogged the silver halides that were not developed in the first developer, limiting development to one layer at a time. A color developer then developed the fogged image, and its exhaustion products reacted with a color coupler to form a dye in the color complementary to the layer's sensitivity. The red-sensitive layer was re-exposed through the base of the film with red light, then redeveloped forming cyan dye. The blue-sensitive layer was re-exposed through the emulsion side of the film with blue light, then redeveloped forming yellow dye. The green-sensitive layer was redeveloped with a developer that chemically fogged it and formed magenta dye.[44] After color development, the metallic silver was converted to silver halide using a bleach solution. The film was then fixed, making these silver halides soluble and leaving only the final dye image. The final steps were to wash the film to remove residual chemicals which might cause deterioration of the dye image, then to dry, cut, and mount the film in slide frames.[44]"
I've highlighted the part that references the fact that the metallic silver is/was converted to silver halide and then the silver halide is/was fixed away, leaving only the dye image.
If silver halides were left in the film after development, then Kodachrome would offer very limited longevity because silver halides are relatively unstable.
If metallic silver was left in the film after development that would be problematic with respect to projection.
As far as the IR absorption of the cyan dye in Kodachrome affecting ICE is concerned, I'm not sure that there is unanimous agreement about how important that is to how well ICE works for a film that has other problems with ICE.
