Why are SOLD/BOUGHT transactions not being included in Investment Performance report?

In the context of my prior post, going after all accounts at once is not particularly different than an individual account.

You have a universe of investments for which you are seeking a performance measure.  That universe can be defined as all your investments, all retirement accounts, any individual account, a group of securities, or any variety of ways.  That defined universe has a size at the start (1-, 3-, 5-, 10 years ago, whenever).  Over time you added to that universe with value that came from outside that universe (cash from a paycheck, transfer from one account outside to an IRA inside, etc.)  Over that same time, you also took value out (paid for kids college, bought a Ferrari, etc.)  At the end of that time, the universe had changed size due to both your additions and withdrawals and internal growth (or shrinkage) of the collection of assets in that defined universe.  Those four lead to the IRR calculation. 

When you exclude the cash from the universe (uncheck the No security (includes cash) box), turning a stock into cash takes value out of the defined universe (the universe gets smaller).  Conversely buying a new security (one that is in the defined universe) adds value (size) to the universe.  The cash was outside, now that value is inside.  So excluding cash makes those buys and sells additions to and removals from (investments and returns) that defined universe.  The transactions then appear in the Cash Flow detail for that report. 

(I have wanted to try out that 'universe' language for this type of explanation for some time.  I hope it makes sense.  Just keep in mind how flexible that defined universe might be -- anything from a single security to everything in your entire collection of investments.)   

First Q, not really.  In the first run with the No Security box checked, the cash remains in the whole, in the universe you have defined for the computation.  Buying or selling within the whole does not of itself change the size. 

Is the value excluding the cash more accurate?  I would not judge it that way.  I did your two runs for an all-account universe with (run 1) and without (run 2) the cash for one select period.  The IRR value went up by 0.3% for that period when I excluded the cash (Run 2 shows a better return that Run 1).  Looking at the starting and ending market values, I see the Run 1 data with cash universe a bit bigger on both ends - at the start and at the end - as compared to the run 2 universe.  I effectively started and ended with the same size cash buffer/safety net for that period.  Both universes increased by the same dollar amount (effectively).  The cash did not contribute to any growth. 

Consider a hypothetical to parallel my example -- $1100 grew to $1300 or $1000 grew to $1200.  Both grew by $200.  The first had a $100 cash that did not contribute to growth. The second ignored that component.  The first increased 18.2%; the second 20%.   Which is more 'accurate'?  My answer - neither. Both are equally accurate, but represent different things.    

Now your numbers are undoubtedly different, but I would broadly expect that tendency to be evident.  (Having a cash buffer would normally lessen the highs and the lows).  If you timed the market well, the differential might be much better bigger than my 0.3% differential.  But that advances into a whole range of related questions - how long, how much out, back in when.  

Final Answer?  Run 1 might present your complete picture more fully.  Run 2 might present your security selection performance more accurately.

My kudos to you for experimenting in that fashion.  It is a great way to better understand not only what and how you are doing, but also what Quicken is telling you.