After googling a bit more, I found this paper, Using Wi-Fi for Navigating the Great Indoors. I suppose the algorithm that handles multiple fingerprints, plus compass and accelerometer is what caught Apple's eye.
When a gadget using WiFiSLAM wants to know its location, it analyzes
the signal strengths and unique IDs of all the Wi-Fi networks around
it. That is matched against a reference data set for the area either
accessed over the Internet, or stored on the device. The estimate of
location can be sharpened if a gadget moves slightly, because
WiFiSLAM's algorithms can gather multiple fingerprints. Compass data
and accelerometer signals capturing a person's footsteps are also used
to refine the accuracy of subsequent location fixes as a person moves
WiFiSLAM needs similar data to be gathered in advance inside a
particular building before it can offer location fixes. A person
running another special app must walk around a building a few times,
entering every room at least once. Algorithms originally developed for
robot navigation process the changing pattern of Wi-Fi fingerprints
and footsteps to re-create the path the person covered. That trace is
then manually associated with a map of the place so that WiFiSLAM can
tell a user in that environment where they are.
Edit 2: Also, looks like WifiSlam had a blog that's been removed. However, Google still has it in their cache with some details:
Most recently, WiFiSLAM’s inertial sensor fusion was featured in
Grizzly Analytics. It sparked some excellent e-mail discussion with
Dr. Bruce Krulwich and we’d love to summarize it for you here!
The demo video includes no maps constraints. It is purely accelerometer, gyroscope, and compass.
We are able to get better-than-typical accuracy because we are taking non-traditional pattern-matching approaches to sensor fusion
rather than the conventional “double-integration + Kalman filter”
techniques used traditionally.
We held the phone in front of us, trying to mimic a typical smartphone user who is following a map and walking while looking at
their phone. Nothing super-specific.
Inertial sensor fusion is now enabled by default as of last week’s releases of the entire WiFiSLAM product line: footprint.io, WiFiSLAM
QuickMap, and the Indoor Location SDK. Any user of WiFiSLAM with a
gyroscope-enabled smartphone will receive hybrid positioning that uses
both our Wi-Fi fingerprinting technology combined with our inertial
Edit 3 Grizzly analytics provides map setup details in their recent blog post.
WiFiSlam has released a mobile app that enables any smartphone user to
take a picture of a map of their indoor site, walk around the site a
few times, and have that site work within WiFiSLAM's location
positioning system. This app enables much easier crowd-sourcing of
indoor maps than Google or others have, and would enable iPhone indoor
positioning to spread like wildfire as iPhone fanboys jump to upload
their site maps.
Edit 4 Here's a video from the GeoMeetup (kindly posted by Ragi Burhum) where Joseph Huang of WiFiSLAM presents a talk about the underlying algorithms.