Wikipedia gives 26% Pu-240 (from total Pu inventory) in spent fuel.The obvious answer to the I-131/Xe-131m question is ongoing transient criticalities.
Is there enough pu-240 for significant spontaneous fission?
Can we trust the decay heat calculations, which are based on all fission stopping at SCRAM?
There are some temp spikes iirc in the graphs from the first few days and I don't really buy the "faulty sensors" explanation, because now the sensors seem to be doing just fine, tracking the water injection rates in a most satisfactory manner.
The question could be settled quite easily. But I have this sneaking feeling that TEPCO isn't checking for neutrons, 'cause they don't want to know.