Battle of Battle of Monte Cassino (1943): 20,000 casualties for the Germans 55,000 casualties for the Americans
Battle of Hürtgen Forest (1944): 28,000 casualties for the Germans 33,000 casualties for the Americans
Battle of Villers-Bocage (1944): 8–15 tank for the Germans 23–27 tanks for the Americans
Battle for Brest 4,000 casualties for the Americans 1,000 for the Germans
Battle of the Seelow Heights 30,000 casualties for the Soviets 12,322 casualties for the Germans
Sources: http://en.wikipedia.org/wiki/Category:Conflicts_in_1945 http://en.wikipedia.org/wiki/Category:Conflicts_in_1944
There is a military term "bringing force to a point", the idea is to have lots more of your guys at the place where violence occurs. In the Pacific the Allies had almost complete control over when and where the fight was going to be. The Allies could get a massive superiority of force at the point.
Actually, I wouldn't call the battles you have named (except for Berlin, which was due to Stalin's desire to win the race to take Berlin no matter the casualties) "at the end of WW2".
Take Bocage/Normandy for instance, Germans were fighting with intact, Elite Panzer divisions (i.e the 1st SS), the military situation, while dire, was had not reached collapse level yet.
However, after August 1944, with the destruction of much of the 5th Panzer army at the Falaise Gap and that of Army Group Center during Operation Begaration, the quality of German forces started to drop. German ranks started to being filled by lesser qualitied Volksgrenadier and then Volksturrm militia units as first rate manpower is depleted.
If you look at battles at the very end of WW2 (late 1944/1945), excluding Market Garden, you do start to see Germans taking more casualties (especially when you include prisoners) than the Allies.
For instance, at Aachan you saw around 5000 allied and 10,000 German casualties. At the Ruhr pocket, over 300,000 German soldiers surrendered for far less Allied casualties.
If you look at Rudiger Overman's estimates, the largest amount of casualties (per month) occured in 1945 for Germany:
So to challenge one of your premises, basically I think your sample doesn't really include the battles at the very end of WW2 when German casualties are highest.