Can you guarantee that the completed building will match the performance predicted by the analysis in its day-to-day operations?
I don't think you can completely guarantee the performance because perhaps the weather is different that predicted for the location, or there's more occupancy, more plug load, etc. but this serves as a pretty good estimate - certainly more accurate than what could be down without the software.
When choosing settings for each of the building performance factors, should you always choose the setting that gives the absolute lowest predicted energy use?
Probably not. While energy use is an important factor, there are other factors to consider, cost being a big one. There needs to be a balance of considering energy use, cost, how people will use the building / feel in the building, etc. Always or blindly choosing the lowest predicted energy use may lead to cutting corners in other areas of importance.
How can model-based quantity takeoff improve the design process?
How can designers improve their designs using the information provided by preliminary estimates of the cost of building their design ideas?
What's great about Assemble (or any other software that allows you to see model-based assessments) is that it is very simple to tweak and make alterations. If, in later iterations of a design, more columns are added or certain structural elements are removed, it is not a headache to update and quantify those changes. Alternatively if the price of a building element changes, it is very simple to make the change to the unit cost, and have it reflected immediately in the assessment. It makes the process very easy to share, collaborate on, re-estimate and alter. A more adaptive process means that the final product can be more cost-efficient, more creative, or better suited to the what the designer wants.