The no. 3 reactor yesterday likely experienced another leak from its primary reactor containment (see below). /Update: At 15:55 local time today grey smoke was coming from the south eastern corner of the top of no. 3 (spent fuel pont) and workers were evacuated. The smoke disappeared at 18:00. At 18:30 white smoke came from the west side of no. 2 building./ Work continued on powerline connection and filling spent fuel pools at unit 2, 3 and 4. Unit 5 and 6 seem to be safe for now. No. 1 to 4 are still in trouble. To connect 3 and 4 to outside power will still take several additional days.
Unit 1
A powerline for external power was laid and has been connected to the no. 1 distribution switchboard. Equipment is now getting checked for damage.
Unit 2
A powerline for external power was laid and has been connected to the no. 2 distribution switchboard. Equipment is now getting checked for damage. Seawater is getting filled into the spent fuel pond of no. 2.
Unit 3
The no. 3 spent fuel pool has been filled with seawater. Pressure in the primary containment increased yesterday morning local time (320 kPa as of 11:00 March 20th) and venting, which would release radioactive steam, was prepared. But the pressure then stopped increasing and later decreased (225 kPa as of 22:00 March 20). Willful venting by the operators has not taken place.
But I do believe that the pressure vented itself through a leak in the primary containment. Earlier U.S. tests found drywell flange leaks on General Electric Mark I reactors under high pressure.

A Mark I primary containment with the drywell top in the foreground (source: Wikimedia)
bigger
The drywell top gets bolted onto the containment. The flange between the containment and the drywell top includes a rubber o-ring to prevent leaks. But in tests high pressure inside the containment could lift the top and steam could escape around that o-ring.
Such an "automatic venting" mechanism would explains the earlier hydrogen escapes from the primary containments which eventually destroyed the no 1 and 3 reactor buildings. On Sunday radioactivity at the Daiichi site went up from some 2670 microsievert/hour at 6:30 to some 3350 microsievert at around 14:30 local time. It later came down again and was at around 2360 microsievert at 6:30 on Monday. This correlates with the pressure in no. 3 coming down again after the likely temporary leaking.
Unit 4
Some 160 tons of water were sprayed into the no. 4 spent fuel pool by military firefighters.
Unit 5
Temperatures in the reactor and spend fuel pond has been lowered to below 100 degree centigrade. Electricity is provided via restored emergency generators at unit no. 6. External powerline connection to no. 5 is supposed to be finished today.
Unit 6
Temperatures in the reactor and spend fuel pond has been lowered to below 100 degree centigrade. Two emergency generators are now up and running supplying 5 and 6.
Additional resources:
AllThingsNuclear Union of Concerned Scientists
Bulletin of the Atomic Scientists
Atomic power review blog
Digital Globe Sat Pictures
IAEA Newscenter
NISA Japanese Nuclear Regulator
Japan Atomic Industry Forum (regular updates)
Japanese government press releases in English
Kyodo News Agency
Asahi Shimbun leading Japanese newspaper in English
NHK World TV via Ustream
Status reports for the German Federal Government by the Gesellschaft für Reaktorsicherheit in German language