I did not know PAR3 until now. The claim of PAR3 is made by one paper last year which has only been cited once. In the paper, the author is calling XTR (X transposed region) as PAR3. I have not read through the paper, but I think the evidence described in the abstract is very weak. Correlated copy number changes are not sufficient to define PARs - ordinary segmental duplications are also associated with non-homologous recombinations sometimes. Biologically, a PAR in the middle of chrX is also weird. If there is a single crossover between chrX and chrY in PAR3, would that create a huge hybrid chrX-chrY chromosome? I would take PAR3 with a pile of salt.

On the other hand, XTR is a well-known segmental duplication from chrX to chrY. You can find its approximate boundaries from UCSC self-chain:

https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg19&position=chrX%3A87015838-93805574&hgsid=386334885_AcoZWTXT6wlO6yJPaqfacRL0k0Cl

Note that there are massive inversions and deletions between the chrX/chrY copies (there are papers about it, but I cannot find them now). It is not possible to precisely define the exact boundaries.

EDIT: XTR evolution: <http://www.nature.com/nature/journal/v423/n6942/extref/01722/figures/fig5.pdf>