Beta-1 your point “… and by 100 kts, the fluids will have left the surfaces, leaving only traces behind” is not completely accurate, see parallel thread here
Icing and de-icing. Quite considerable amounts of fluid can remain on the wing; however, your point about them giving no protection is correct.
Your reasoning about unwanted aerodynamic changes is partially correct; previously the industry had considerable concerns about this, but research showed that any reduction in aircraft performance was small (737). However, recent tests conducted by some ‘regional aircraft’ manufacturers suggest that there could be noticeable effects on climb performance, particularly if large quantities of fluid remained on the wing due to incorrect application etc. Furthermore, several turboprops have significant changes to operating procedures when using de-icing fluids; changes to take-off trim setting and take-off speeds; these were due to aircraft control problems.
In addition to the misuse of fluids, one reason for these problems could be the use of fluids with high ‘dynamic viscosity’ – the viscosity is proportional to airflow. As speed increases, the fluid will flow more easily, but if fluid moves into lower speed areas (flap gap, trailing edge, and control gaps), viscosity increases and it tends to stick and thus not all of the fluid comes off.