Speaking from a mechanical/materials/structural engineering perspective, it stands to reason that making a significant alteration in a highly integrated system, such as adding a carbon fork to a steel bike frame, can have unintended consequences. I certainly don't rule out the possibility that changing to a carbon fork can move the normal stresses induced on the head tube area of the frame from a value below steel's fatigue limit to a value that is above its fatigue limit.
This has been covered before in fairly recent threads, of common bikeframe metal alloys, only steel and titanium have fatigue limits, anything else can be guaranteed to fail eventually under regular use. Stell and titanium will not, as long as they are not subjected to many cycles of repetitive stress above a given alloy's respective fatigue limit.
Composites (e.g. "carbon") have their own issue, which is relative intolerance to surface damage. Surface damage is a given with any article that gets used in the real world. A protective coating would help, and for all I know about carbon frames/forks such coatings may in fact be used, but there would be a weight addition without a stiffness or yield strength benefit.