YES It Does !
... I realized the perfect place to test whether housing location always matters is a power plant. I think the power output is completely independent of whether or not a plant worker is on the job or wandering around the island. From what I remember, the power output is equal to the number of workers times power per worker. It only changes when you gain or lose workers, it does not change throughout time as workers show up or leave the plant. Otherwise power output would constantly fluctuate and you would have random blackouts when by chance all the plant operators were out of the office at the same time. This to me argues that there is at least one place [Then it is NOT the perfect place for a test.] where it doesn't matter where the employee lives. I think caddet definitely could be right in that absent teachers might equal slower students, or absent maids equal unhappy hotel guests. I don't know how you'd test that theory, though, since there are so many other variables. ...
Bad logic, poor test. If the building is unique, it is not a proper test for formulating a general rule. This fallacious line of reasoning led to the Urban Legend that
indoor work does not require actual presence of the workers at any time. Mjara's guess is the correct answer. It is very difficult to measure the actual output of the power plant to determine if there is a fluctuation -- particularly to determine if there is a gradual downgrade if no Engineer actually enters the building over a long period of time. We know that the Engineers improve their skill levels only during actual time on the job (without the cheat "Learning With Larry"), so actual presence has something to do with it. Let's examine the issue in a bit more detail.
All building which allow units to enter (that is, all except the decorations such as the Statue) are production buildings which produce or pass through product or units at a variable speed. Some have to do with activity outside the building - "outside animations" provide interest and amusement for the player. Most have to do with something going on inside the building - "inside activity" which is invisible to the player. In fact, it is rather difficult (not intuitive) for the player to realize when an employee is actually in the building. T2 (the pirates) corrected this by showing the avatar in the building display in two different modes.
There are some basic types of buildings:
- Initial Production
- Farms & Mines, etc.
- Electric Power Plant, etc.
- Product Change
- Unit Needs Gage Recharge
- Housing & service buildings
- Gameplay Effects Only
- Diplomatic Ministry, Bank, etc.
.
A rather simple level of observation has confirmed that factories move product from the input queue to the output queue faster when employees are actually in the building. With respect to the service buildings, it has been much more comfortable to accept the 'Urban Legend' than to take the trouble to confirm the relationship of movement of units through the building with the actual presence of the workers.
To assume that there is a great divergence between the importance of actual presence between outdoor (animation) work and indoor (invisible) work is to assume that the developers were a bunch of blithering idiots. Why would they not use the same standard lines of code for both situations?