Typically, modern-day code-breakers focus (if not over-focus) on the transcription and computer analyses. However, people are sometimes motivated by quite different things from pure security - the psychology is at least as important. Pigpen is easy because you can decrypt it very fast (an arrow-based pigpen would be at least as quick to read as a 'proper' one), and perhaps this is what Olindo Romano wanted. And it seems likely to me that he thought/thinks he's cleverer than all the people around him (whether that's true or not).
Of course, this is the kind of approach I have used when looking at the Voynich Manuscript, so it should come as no surprise that this is how I look at things. I wonder: if the Italian "mathematicians" who have deciphered this cipher plotted out the letters they have found on 3x3 arrow-pigpen grids, what would they find?