I would like to clarify why I asked you to take more pictures. I've found in my files some notes I took many years ago, when I worked on a similar unit. At that time, I've made a very partial schematic (not publishable) and from the schematic I have, I am quite confident that you will not fix this problem if you replace the 2SA798. The two new pictures will help me to revise and complete the schematic I have.
It is my believe that if you turn the dual transistors the other way, the bevel facing the opposite direction, your offset will be approximately the same magnitude and the same polarity. Actually, your offset is certainly positive. If your offset was caused only by a mismatch in those dual transistors, you would have a new offset of the same magnitude but of opposite polarity, by turning the dual transistors the other way.
So, if you turn the dual transistors and you have the same offset polarity, it is a waste of time and money to replace those dual transistors. Your offset, like in any other SA-200, is caused by an error of conception but it can be corrected.
From the schematic I have, and the colour bars I can read from your pictures, I would say that you can correct the offset if you replace resistors R613 and R614 (actually 470 Ohms) by ~130 Ohms resistors. You can connect another resistor in parallel with the actual resistor, with small jumpers, starting with a 1 kOhms, and progressively reducing this second resistor until the offset is near 0mV. If the parallel resistor is too low, the offset will become negative. The optimum parallel resistor should be close to 180 Ohms. Once the optimum bypass resistor is found, you can solder it in parallel with the actual resistor, and your offset will stay very low for years to come.