Your answer matches my intuition, that the blanket spread out would do a better job of insulating, but I don't fully understand your math. Am I correct in reading you as calculating the total resistance of the floor? What is the equation you're using in the folded blanket example? I'm just not sure how you're calculating parallel resistance there.

And is that really the best way to calculate it? It doesn't seem obvious that changing insulation on one part of the floor will affect the cooling effect of the other parts, but isn't that the case for resistance, at least where the two halves are treated as parallel paths?

My intuition was to think of the problem in terms of rates, and ask how the blanket would affect the rate of heat flow. If an insulator decreases heat flow proportionally, i.e. by x%, then doubling the blanket would be less effective: The first layer of blanket would decrease the rate by x%, and then the second layer would decrease that lowered rate by x%. I suspect this is how insulation works.

If, however, an insulator reduced the rate by a fixed amount rather than a percent, then doubling up would be no different.

Finally, if insulation increased non-linearly, so that 2x the insulation produced >2x the reduction in rate, doubling could reduced total heat flow. The situation I've come up with where this might be the case is one where the blankets have a checkerboard pattern of open and closed spaces, so that when doubled they are fully closed. This could create a super-linear increase in insulation if a large obstruction to airflow is more effective than several small obstructions that sum to the same area.