You are right.
The dots are phasing dots. If the primary has a signal such that the end of the winding with the dot is positive, the secondary end with the dot will also be positive at that instant. Here's a tip: it looks like Hammond boogered Fender's original spec; the brown wire should be connected to the 6L6 which is fed from the 82k half of the phase inverter. Original Fender layouts show the blue wire at this point.
The catch is there's no single right connection for
all feedback circuits (compare feedback injection between the AA864 Bassman and AB165 Bassman). We can only determine right if we select which circuit we're discussing.
For the AA864 Bassman, which is typical of most blackface amps, the OT should not invert polarity for the feedback to remain negative. The phase inverter section with the 82k resistor will be our reference point. The 6L6 inverts the signal present at the 82k plate, and then the signal goes to the OT. From the secondary of the OT, the feedback signal is applied to the bottom of the tail resistor, which is effectively the cathode of the 82k section. A signal applied to the cathode of a tube causes an output of the same polarity, so for the feedback to be negative, we don't want to OT to invert polarity.
The problem is it looks like Hammond did Fender's colors backwards. The blue wire in the Fender OT should be same-polarity as the ungrounded end of the secondary. On the Hammond data sheet, the blue wire is same-polarity as the grounded end of the secondary (black wire). Therefore, the brown wire should attach to the plate of the 6L6 fed from the 82k section of the phase inverter.
Now, someone who has actually used the OT in question oughta chime in and verify if this is actually correct!
