The “Sailboat Island” gets its name from a Poincaré Surface-of-Section (SOS) plot of potentially stable S-type family of orbits in the Pluto-Charon system. From NASA Ames' blog post titled Playing Marbles at Pluto. Looking at the Dynamic Dust Environment. Generators, Sweepers, and Sweet-Spots by Kimberly Ennico:
Othon Winter (UNESP Brazil) spoke about “On the Relevance of the
Sailboat Island for the New Horizons Mission.” In investigating where
particles would find stable orbits, their modeling predicted a region
where there was a cluster of orbits characterized by high eccentricity
(e= 0.2 to 0.8) and located around 0.6 Pluto-Charon semi-major axis
(i.e. between Pluto and Charon). They nicknamed it “Sailboat Island’
because on a eccentricity vs. distance from Pluto plot it looked like
a sailboat. This population of “stable orbits” had not been predicted
from previous work.

The figure above is taken from Giuliatti Winter et al 2010 where they
describe a family orbits called S-type that are stable. The plots are
in d vs. e. where d, on the x axis is the Pluto-centric semi-major
axis (how far from the Pluto barycenter) and e, on the y axis is the
eccentricity. The “white” areas are orbit solution that were found to
be stable. Area ‘1’ is the “Sailboat Island” described in the talk.
Left are prograde (inclination=0) orbits, right are retrograde
(inclination=180 degrees) orbits.
From A PECULIAR STABLE REGION AROUND PLUTO AND ITS ROLE ON THE NEW HORIZONS MISSION TRAJECTORY (PDF), whose list of authors, among others, includes already mentioned Othon C. and Silvia M. Giuliatti Winter, UNESP São Paulo State University, Brazil:

Figure 1: Diagram of $a$ [semi-major axis] versus $e$ [eccentricity]. The small stable region is shown
in black. The nominal parameters of the particles are: $ω,= 0$ and $Ƭ
> = 0$, where $ω$ is the argument of pericentre and $Ƭ$ is the epoch of the pericentre.

Figure 2: A sample of Poincaré surface of sections for six different
values of the Jacobi Constant. In each case are shown only those
points associated to the periodic and quasi-periodic orbits of the
family “BD” (Broucke, 1968), which corresponds to the small region
(Fig. 1).

Figure 3: The set of periodic orbits, in the synodic frame, for the
different values of $C_J$ as shown in Figure 2. The barycentre is
located at 0.
New Horizons' nominal trajectory passes near this region of “Sailboat Island” trajectories, the closest at about 1650 km. It is still unclear if New Horizons will have to adjust its trajectory to avoid any possible debris in its path, but with observations made so far, all seems clear. Alan Stern, New Horizons PI, describes possible alternative flyby trajectories in this article.
Some additional reading: