I suspect the issue is that there are two few cases to extrapolate to smaller geographies. I just did a quick check in Factfinder for Cambridge, MA, population around 105,000. The only instances that came up were 1 year tables from the 2008 and 2009 ACS. In both cases there was too small a sample and the tables were suppressed. This is a bit surprising since the Census recently published a paper on grandchildren living with grandparents, not exactly a rare household arrangement. There might be relevant tables available through the 2010 Census.