I've seen discussion that the heat lost from the relatively warm ocean water when not covered by ice exceeds the additional heat absorbed by low albedo, suggesting that lack of ice cover is actually a net negative feedback.
I have no personal calculations as to the magnitude of either effect, but I am skeptical of positive feedback loops, as they are inherently unstable. Note that we get significant ice loss every summer, followed by rapid gains in ice, which seems to cast doubt on the significance of any positive feedback.
Open water loses heat much more rapidly than ice-covered water, with the rate dependent on temperature, wind speed, etc. One study found open water heat flux was about 250 W/m2 vs 75 for ice pack in March/April (http://onlinelibrary.wiley.com/doi/10.1029/94JC02485/abstrac...). I didn't find references to give averages or graphs over a year.
However long the ocean remains open, it radiates excess heat at an increased rate vs ice pack - a negative feedback of open water, limited by the area and duration of open water.