My papers and those of Jan-Erik Solheim et al predict a significant cooling over Solar Cycle 24 relative to Solar Cycle 23. Solheim's model predicts that Solar Cycle 24, for the northern hemisphere, will be 0.9º C cooler than Solar Cycle 23. It hasn't cooled yet and we are three and a half years into the current cycle. The longer the temperature stays where it is, the more cooling has to come over the rest of the cycle for the predicted average reduction to occur.

So when will it cool? As Nir Shaviv and others have noted, the biggest calorimeter on the plant is the oceans. My work on sea level response to solar activity (link) found that the breakover between sea level rise and sea level fall is a sunspot amplitude of 40:
Image
As this graph from SIDC shows, the current solar amplitude is about 60 in the run-up to solar maximum, expected in May 2013:
Image

The two remaining variables in our quest are the timing of the sunspot number fall below 40 and the length of Solar Cycle 24. So far, Solar Cycle 24 is shaping up almost exactly like Solar Cycle 5, the first half of the Dalton Minimum:
Image
The heliospheric current sheet tilt angle has reached the level at which solar maximum occurs. It usually spends a year at this level before heading back down again:
Image
Similarly, the solar polar field strength (from the Wilcox Solar Observatory) suggest that solar maximum may be up to a year away:
Image
Notwithstanding that solar maximum, as predicted from heliocentric current sheet tilt angle and solar polar field strength, is still a little way off, if Solar Cycle 24 continues to shape up like Solar Cycle 5, sunspot amplitude will fall below 40 from mid-2013. Altrock's green corona emissions diagramme (link) suggests that Solar Cycle 24 will be 17 years long, ending in 2026. That leaves twelve and a half years of cooling from mid-2013.

From all that, for Solheim's predicted temperature decline of 0.9º C over the whole of Solar Cycle 24 to be achieved, the decline from mid-2013 will be 1.2º C on average over the then remaining twelve and a half years of the cycle. No doubt the cooling will be back-loaded, making the further decline predicted over Solar Cycle 25 relative to Solar Cycle 24 more readily achievable.