If you carefully look at your composite zona.jpg, you see that the main eastern coordinates have a gap at the 30° meridian, and the latitudes have a sharp bend there. There is a secondary coordinate line on the top, that follows the western CRS on the eastern part, and the eastern CRS on the western. Same for the latitudes.
The EPSG definitions of 3-degree Gauss-Krueger have central meridians at 24/27/30/33°, but the imprinted CRS does not fit to these: You would have the false Easting of 500000 (EPSG:2585) or 10500000 (EPSG:2526) at 30°, but instead you have a gap there.
Instead , the 6-degree Gauss-Krueger CRS fit better for the western and eastern half:
Pulkovo 1942 / Gauss-Kruger zone 5 : +proj=tmerc +lat_0=0 +lon_0=27 +k=1 +x_0=5500000 +y_0=0 +ellps=krass +towgs84=23.92,-141.27,-80.9,-0,0.35,0.82,-0.12 +units=m +no_defs
Pulkovo 1942 / Gauss-Kruger zone 6 : +proj=tmerc +lat_0=0 +lon_0=33 +k=1 +x_0=6500000 +y_0=0 +ellps=krass +towgs84=23.92,-141.27,-80.9,-0,0.35,0.82,-0.12 +units=m +no_defs
This seems to be similar to UTM or EPSG:28465/28466, except for the leading zone number.
So you can georeference your single maps to the respective zone CRS, and reproject one half to the CRS of the other in a second step. Or use the inner coordinates for one half, and the outer coordinates for the other.
Georeferencing by using the degree coordinates as EPSG:4284 (Pulkovo 1942) is another option, but might invoke some disturbances, since the map is projected to meters, not degrees.