The Face in Plant 17v

[VViews]

Hi Koen Gh,
IMO, the first stop here is internal comparison.
The profile view is evidently not the Voynich artist's preferred choice for the representation of faces. Quite the contrary: there are hundreds of faces in the Voynich manuscript and not a single one of them is shown in profile.
If it was a profile, then this would imply that this particular image is of a wholly different inspiration (copied from a totally different source?) than all of the rest of the manuscript, or that it represents a person of a completely different nature or status from all the other people in the manuscript. And although its fun to speculate, it's not the sort of testable thing you're after.
Profile views versus other angles in human representation happens to be something I actually researched in an academic context. It is a very culture-laden stylistic element, a type of representation that doesn't just "happen" by chance. There are whole civilizations that simply don't "do" profile, while for others, it is the go-to style. It is plain to see that profile views are not a part of the Voynich artist's habit. Why would one suddenly appear here?
Also, the fact that "tendrils in other manuscripts generally don't look like this" doesn't heighten the probability that this is a face. To strengthen your hypothesis you'd need to find other examples of profiles made of plants.  AFAIK there's not much before Arcimboldo, and that's a century too late.

I am not denying that artists have hidden faces in works of art. It is rare, especially before the modern period. The first ever known instance of this is the rather well-known case of a face in profile in a cloud in Giotto's mural in the Basilica of St Francis of Assisi (1290).
But again, to my point above: profile views are habitual in Giotto's work, (as in Arcimboldo's btw) but not in the Voynich artist's.  So even this example actually doesn't strengthen your tendrilface hypothesis.
For fun: Can you spot the profile in the cloud?


Finally, I want to offer this map as a counter example: the lower landmass could be seen as an upward looking profile, with the same rounded features as your "tendrilface". Except it's not a face, just a landmass that sort of looks like one. 

Personally, (I mean no offence) but I think "tendrilface" belongs to the pareidolia category.