Implementation shortfall computed differently

Implementation shortfall is defined as “the difference between the money return on a notional or paper portfolio and the actual portfolio return” (from the glossary). However, the solution to question 10.B. of reading 44, on page 62 of volume 6, computes the implementation shortfall in a completely different way. It uses the quotation midpoint as the benchmark price - whereas by definition the return on a notional portfolio can never use a midpoint as a starting comparison point since nobody can buy or sell shares at a midpoint quotation price. One either sells at bid or buys at ask. How come the CFAI official text presents a solution that contradicts the definition of a key concept introduced both in the main body of the reading and in the glossary?

I really can’t comment on this as I can’t figure it out, but I do notice that there is a lot of conflicting information in the readings. If they give us a question with bid-ask price, I’m just going to use the midpoint and take that as a given.

Not sure either, but I will use the BP as the first price when the manager decides he wants to get in on the stock.

If there have been no trades in the last 1 hour of the day what do you do for a price? Some people would use the bid-ask mid point as a fair price.