I'm using the 2445B image, because the measuring errors are satisfactory for me without calibration. Your images also operate normally, but the calibration data does not match to my hardware - error of measurement is 15...20% in my scope.
Looks like all analog Tek 24xxB scope's SRAM image are compatible each other, never expected that, great finding.

I don't know the part no of this BBSRAM, originated from an old RICOH photocopying machine, was located on main control board PWB no. A1535112. The EPSON chip is a 256K-BIT SRAM with Extremely Low Standby Current. It's like a Dallas DS1225 without cover and with larger battery
. Address lines A13, A14 are grounded and disconected from scopes board. Did not work properly as they were connected.
Yeah, I was curious when seeing the pin 1 & 26 mod job, no wonder.
About this BBSRAM at Tek 24x5B scope, an interesting finding, if you check the scope schematic, the pin 1 and 26 are connected, while at the BBSRAM DS1225Y chip, those are NC pins.

Probably they were reserving that A13 & A14 lines for optional expansion when the scope was designed back then. Your problem with that grounded 2 extra unused address lines is confirming that, both pins are supposed to be pulled high or as in no connections as DS1225Y does, pulling both down definitely will confuse the CPU.
I am curious how it goes with FM28V020. It's hard to get !
Yeah, its hard to find, since this F-RAM is not as common as other like EEPROM, just a warning, never buy these kind of rare memory type from Chinese online store like at ebay or aliexpress, trust me, 99.99% are fake.
Ordered my FM28V020 directly at Cypress Semi's online store, even though not that cheap @$21.69 a pop

, they're providing "free" shipping, so its worth compared to buying the pesky battery based DS1225-AD (DS1225Y replacement) through Maxim's distributors ranging about $17 to $20 price range but with crazy s/h charge.

Btw, I've been reading intensely the 256K FM28V020 F-RAM datasheet and comparing with DS1225Y lately, and so far it is quite promising. With the 4 times bigger capacity, I'm going to route the two extra unused address lines into a dip switch with 2 positions, so basically I could manually switch between the 4 x 64K memory space, enough for experimenting with different calibration values, or easy hacking/tweaking around the SRAM's content if needed

.
For physical compatibility, since FM28V020 is SOIC 28, its obvious I need a custom breakout PCB for SOIC 28 to DIP 28 and with a dip switch placed at the PCB.
Btw, thanks for the U800 heat sinking photos.
