Unofficial Fallout 4 Patch

Before going into the details, I need to know a few things:

(1) How frequent is that issue in your game ?
(2) Does it only happen to Sanctuary ?
(3) How far did you progress in your current game ?

1 - Happened twice so far
2 - Any settlement, before sanctuary was tenpines bluff, saw it happening on the castle on other save too. And it's not just the water, sometimes it reports missing beds too.
3 - Finished main story and a good portion of side quests

Until my last post, we had focussed on threading issues to get this issue solved. All of the related fixes made the issue occur more or less likely but we eventually came to a point where the threading issues were fixed but the problem remained. During the last year, I have invested much time in improving the checks for workshops being loaded and in modifiying the scripts to stop recalculating workshop resource rating values - or at least to suspend them temporarily - when there's sufficient evidence to assume that a workshop has partially unloaded. This further improved the situation, i.e. it made the bug occur even less likely, at least for smaller workshops. But it also became obvious that we cannot permanently fix the issue. Last week, I posted the following summary on the Nexus:

Settlement stats are not just affected by one bug. Looking at the issue as an outsider, with knowledge limited to in-game experience, one is likely inclined to think that there must be something wrong with the calculation that prints the data on the pip-boy screen, and that this should be pretty easy to fix. Well, if that was the case, it would have been one of the easiest fixes ever.

The sad truth however is, that there is nothing wrong with the calculations (it's actually not just one calculation but several of them, in various parts of the code of two scripts, but there's nothing wrong with them either).. What is wrong instead is the workflow: the functions that operate on the resource values are supplied with wrong input data, either with data from different workshops or with invalid data. The issues fall in two categories:

(1) Threading issues:, see:: https://www.creationkit.com/fallout4/index.php?title=Threading_Notes_(Papyrus) :
Those issues all have in common that their results are unpredictable, Therefore, you also can't track them the other way around, because repeatable deviations from expected results do unfortunately not mean that the error occured in repeatable conditions. The only way to spot them is by reading through the scripts line by line. If you check the changelog for anything related to "threading", "thread locks" or "edit locks", you'll find around two dozens of fixes, most of which contributed to making the workshop reource statistics bug less likely to occur (some more, others less).

(2) Issues with unloaded workshops:
Because the game stores all its workshop stuff on cell data, objects that are in unloaded cells while the calculations are running will not be found and their contribution to the workshop resource values will be missing from the grand total. This issue cannot be fixed because papyrus is missing a function to check whether all cells of a location are loaded (so this is essentially an engine issue). The existing function to check whether a location is loaded returns true (i.e. it considers the location as loaded) if any of its cells is loaded, so it would still return true if only one cell is loaded. Needless to say that this is completely inappropriate for checking workshop locations. We improved this by adding additional checks, but this only works for workshops that fit into a square of 3x3 cells. Larger workshops are still affected by this bug and there's nothing we can do about it.
It's worth noting here that the 3x3 cell condition does not refer to a workshop's actual size but to the number of cells that are to some part whithin the buildable area. Jamaica Plains, for example, has the size of roughly one cell but it is centerd on a point where four cells connect, so it occupies a 2x2 cell square. If you make a workshop that fills one cell and extends for only a small distance into all neighbouring cells, it technically occupies a 3x3 cell square. Now, if you make another workshop that fills two connected cells and extends for a small distance into all neighbouring cells, it's still a small workshop but it does no longer fulfill the 3x3 cells criterion. Needless to say that you can also make things significantly worse by running a mod that increases the workshop sizes.

To put it short: there's nothing else we can do.