The first Sony Aibo had the same problem - interesting core technology, but a closed architecture without a viable host communication link, and a concept that there might be a market for a $2500 robotic pet. I owned one of the first Aibo's, but sold it shortly thereafter on ebay because it was basically useless. Sony later added Wifi and opened the platform to developers, and Aibo had a very good run as the basis for 4-legged Robocup soccer and numerous robot development platforms. It ultimately sold more than 150,000 units at $1700+, and might have continued life if Sony hadn't gone through restructuring and cut a lot of programs.
The Pleo was likewise based on the same flawed vision, but they didn't make the transition in time. Closed architecture and lack of wireless communications is a dead end for this type of platform, but a wireless programmable robot with camera and 14 servos for $$350 would have gained a lot of support from the developer community and ultimately could have done quite well.