gatt writes:

Those sensations are very important. Knowing how to interpret them
(and how to avoid misinterpreting them) is especially important.

Then why must they be ignored for safe IFR flight?

"There are three sources of actual 'feel' that are very important to the
pilot. One is the pilot's own body as it responds to the forces of
acceleration. The 'G' loads imposed on the airframe are also felt by the
pilot. Centripetal accelerations for the pilot down into the seat or
raise the pilot against the seat belt. Radial accelerations, as they
produce slips or skids of the airframe, shift the pilot from side to
side in the seat. These forces need not be strong, only perceptible by
the pilot to be useful. An accomplished pilot who had excellent 'feel'
for the airplane will be able to detect even the minutest change.

How do pilots of RC models and UAVs manage to fly, given that they do not have
these sensations?

I understand why so many pilots without IFR training last only a few minutes
in IMC before they spin out of control, if they have such an overwhelming
dependence on relatively unreliable physical sensations.