Looks like it was a combo between the reduction of fusealge drag and cooling drag that gave the D9 the advantage.
See here for 'Rumpf' and 'Kuhlung' row at the drag analysis, where fuselage is 10% less draggy on the D9, and cooling drag is lower by 50%. Weapon set-up is also less draggy - no wonders here, the D9 lacked two cannons and their associated drag.
Though I do not quite know what the two different figures at Widerstandsflächen exactly mean it is quite interesting to compare the various models and their differences.
The A and D models have the same wing so the first numbers are identical (0,1647) but the second figures of the Ds differ (34,0/34,0/37,1/37,3).
What does it mean?
The wings of the Ta 152s (34,8/36,0/38,6). Why do the numbers for the wings of C and E, though identical differ?
The tail units of the Doras and Ta 152 models have the same first figures (0,470), which means they must have the uniform or Ta 152 vertical stabilizer.
The second figures differ in all, except for the A models. (9,1/9,1/10,6/10,7/9,3/9,6/9,4). Why is that?
The second figures of tht fuselages and KA (gun installation?) of D-9 and D-12 differ slightly (31,4/31,6). I think this is because of the greater drag of the cowl mounted guns. Ta 152 E and H differ probably because of the recon equipment in the E uselage.
Interesting is the higher drag of the air intake for the supercharger of the Jumo 213E/F compared to the Ta 152C's DB603L.