In the BeerSmith program, BIAB is defined as only the original method developed in Australia which is a full volume mash. So avoid using a BIAB mash profile. For all intents and purposes, your filter bag is really just the false bottom or screen at the bottom of the lauter tun. Since you are sparging, you will need to use a standard infusion with batch sparge type of a profile.
The next issue you will run up against is how BeerSmith calculates the needed volumes. Since the program calculates backwards from the final volume into the fermenter, it needs to have one variable which floats in order to balance out the water input and this variable is the sparge water. To do this, it defines the strike water as the ratio given of water to the grain bill plus any dead space. Note that it uses the ratio of water to grain and not a fixed value.
If you are comfortable with this, then start by entering your 12 liters in your initial mash profile and give the program a typical grain bill size. It only uses this figure to calculate your water to grain ratio. When applied to a recipe, the actual strike water volume will vary based upon the size of the grain bill you use. If you are OK with this you are set. If you want to maintain a strict 12 liter initial infusion, you will have to adjust the mash profile in each recipe after you apply it to the recipe.
The one thing that your method as described does not take into account is the amount of water absorbed by the grains. So the variable you are floating in your process is the pre-boil volume (which then dictates the subsequent downstream volumes. I am not sure how you manage that variation other than topping off when you run short. The purpose of the program is to help define exactly how much water you will need from the beginning. In the end it is your process and it it is working correctly for you, then I would create the profile as above, make an adjustment when applied to the recipe to give you a 12 liter infusion volume, and then use the sparge volume to either adjust how much water you use for your sparge or stick with your 7 liters and use it to help anticipate if you will need to top off.
In terms of using a separate pot and batch sparging there that has no bearing on the program. All it knows is that you did a batch sparge and the results of that will be reflected in your efficiency numbers.
Without reference as to the changes, I am guessing that your salt additions are changing due to the difference in required water (due to the change in grain absorption).
There are a couple of thoughts with regard to addition of water agents. the first is to take all of the salts and add them to the strike water in order to help with control of the pH during the mash. The second is to divide them up based upon volume and treat all the water equally. Either works fine. Brad has a blog post on how to enter in your water and water agents in BS3 and there are several threads you can search which describe the process. I do believe that once you figure out how you want to handle the mash profile, this will become a bit easier for you to tackle.
