Have any studies been conducted to decide the 'print angle' (anamorphic projection) of road signage?
Are there any other expectations involved?
I do not think there is an known answer to this question.
There is research on other horizontal signing as speed countermeasures in curves that attempts to use pavement marking size and spacing to lead drivers to the appropriate speed. I am unaware of research that addresses the use of arrow dimensions for the same purpose.
Starting with general information and leading toward more specific research, here are some references that might help guide you toward your own answer.
Human Factors Guidelines for Road Systems. See Chapter 6 on curves, speed selection in curves, and countermeasures.
Toolbox of Countermeasures for Rural Two-Lane Curves. See pages 44 through 45
Those documents point to two experiments using arrows as pavement markings.
The most recent has two sections of interest.
This research paper describes an experiment evaluating the effect of an arrow preceding the entry into a curve on vehicle speed. They do not describe how they decided on the dimensions of the arrow or the word SLOW as shown in Figure 2 in the paper.
That paper referenced this earlier research. It lists the variables to consider when placing and sizing pavement markings. It is applied to stripes, rather than arrows, preceding a curve but should provide a good starting point for solving the problem.
As always, if I've missed something, please amend the answer to include it.