Hello!
Yesterday I was carrying out more bombing tests, and I realized that actually with the Lofte BS the manual targeting is much more accurate than the auto targeting. Well... really we could say:
actually the manual targeting is the only accurate method for level bombing. 
This has a problem: the player should do all the calculations related with 'parabolic motion' and the BS elevation angle by himself.
I also tested the manual targeting with the OKPB-1 bombsight. And, surprisingly, I realized that the in-game calculation indicates an elevation angle greater than the calculated manually. The difference is about 3ยบ, regarding of the altitude and the OAT, and that is sufficient to produce the observed longer drops with the russian bombers.
Thus:
The intended aiming method of both types of bombsights (Lofte/Norden and OKPB-1) don't work as it should do.
Other important matters about manual targeting with both bombsights:
- The TAS must be calculated taking in account both the Indicated Altitude and the OAT at such altitude (remember: the OAT depends of the chosen map).
-The altitude to use for the BS' elevation angle calculation must be the True Altitude: it is found from the Indicated Altitude affected by OAT and then subtracting the Target Altitude.
All indicates that now the atmosphere model is more accurate and more consistent with the laws of Physics, and the aircrafts as well as bombs exhibit a more complex but predictable behavior according with a given combination of external conditions: altitude, OAT, wind, etc. Almost like IRL...
But perhaps something was missed about the bombsights along the succesive improvements, since the patch 4.10 until today.
I'll do more tests with different maps and with American and Japanese bombers. In the best case, such tests should confirm what was observed with the He-111 and the Pe-8.