Latency has nothing to do with dizzy cap size, which are the same diameter between the 2.8 and the 4.3 BTW, but the latency is a value of how long it takes the ICM to have an output to teh coil from when the signal is input to it. In otherwords, the delay between incoming and outgoing signals. This is module dependent, and can have some effect if the wrong values are used, but probably not as much as you are thinking.

The TPS readings between both the 2.8 and 4.3 throttle bodies are the same, at the very least they are both linear 0 to 5 volt signals, What can change however is HOW MUCH the throttle is opened for the same airflow between each throttle body, since one is larger than the other. This sometimes means that some throttle position values that enable or disable functions might need to be changed, such as when certain tables are used or when idle parameters are entered and exited.

I'm not a fan of using values from other masks, some masks will have different calculations for the same values, and blindly substituting values may cause more problems than it solves. Obviously displacement, cylinder size, injector size etc can all be substuted because these are based on physical attributes of the engine itself. Going back to the spark latency table I would probably copy that one over, since you will be using the 2.8 ICM, though the values may be close to start with anyway.

Set the initial spark advance to what the actual initial spark advance is. I have used the bias table in the past to get the actual timing to be closer (or exactly) what is seen in the SA table.