The 6SN7 will barely drive a 300b, which is a very difficult load. I have not worked with 845 tubes nor will I given the high voltages and heat involved, but I did look at the data sheet. Yes, it takes more negative grid voltage to bias one than a 300b, but I have no idea of how capacitive a load an 845 is, and I really don't have time to research it. Ralph is right in that a 6SN7 is adequate to drive many tubes. It makes a wonderful driver for any of the usual octal pentodes like kt88 or el34, or a 6V6 or 6L6. Works great. But a 300b taxes a 6SN7. It works as in the original Reichert amp, but if you want to unlock the unlimited dynamics of the 300b you need a much better driver setup. The reason the original Dynaco ST70 was a marignal amp was it had a wimpy driver section that would clip before the output tubes, or about the same time, plus it had a very mediocre power supply. It was a pleasant amp and a zillion of them sold. But if the driver and power supply were better it would have been a much better amp. The output transformers are not world class, but they are quite good and capable of more than the rest of the amp could provide.
You can also improve the performance of any driver tube by changing the way it is loaded. For example, a CCS will give far better drive capability than just typical RC coupling. But CCS circuits do have some coloration. Not so bad in a pentode amp, but very noticeable in a DHT amp. The other ways are to add inductance to the plate of the driver with either LC (choke cap) coupling or full on interstage transformer coupling. The inductance the driver tube plate sees "helps" it out to drive the 300b. There are differences in the way all of these approaches sound, but all are better than just a plate load resistor and coupling cap. I have explored all of them thoroughly for a year with the 300b project and found a solution that Lynn and I like and sounds the most musical.