these are the wrong way around, followup commits didn't notice and left it like this

...because the properties are neither gui nor sim data

(and then others don't sort either D1058)

Missing *
The order of functions alreadyi s completely random in this file anyhow, so not really a regression, but functions should be sorted, first input, then processing, then output (https://en.wikipedia.org/wiki/IPO_model) and the shorter functions first.

So it's an empty string if undefined and an object with two properties otherwise?

should use tabs instead of spaces

The reader might want to lookup the code if it's needed to distinguish empty string from undefined

Why isn't g_SelectedPanel grouped with g_SelectedChart as it is below?
Why are there two variables here while having an object with two properties elsewhere?

rP20684,
yet the * is missing

there lies a deeper duplication issue here: the way tabs in the summary screen are handled is fundamentally different from what we do in the credits/options and as I didn't want to mess with that in rP20684, it isn't nuked there. A proper fix should nuke the duplication and then the global would move to another file and become an index instead of a string, thus fixing both issues.

Dialog toggle hotkeys should be used for opening or closing dialogs, but not for switching the current page away (as dialog actions are reversible, but page switches are not), so IMO the page switch changes are better off with a separate hotkey. For example we also dont want someone to close the gamesetup page after having setup the settings (for sometimes half an hour) and then pressing escape (but with shift+esape it would be ok, similar to delete without confirmation hotkey).

For the replay menu, it is also a "recoverable" action. Mostly the page should be opened as a dialog, but the dialog looks ugly, as those outer 20-ish margin pixels are mostly black, while ingame they show the map and units.
But closing the patch after having finished a game with the close hotkeys still seems wrong to me, it should be consistent for all pages.
So perhaps this can be improved:

explicitResume is meant to be a boolean, undefined if you will, but not 0.
If the return value is not defined, undefined, or otherwise falsy, it doesnt have to be passed to begin with.

The 0 return value from PopGuiPageCB seems to have just been used as a placeholder because anything must be returned in the code prior to rP22676 if the page was opened with a callback argument IIRC.

The 0 seems to oringate at rP17422 but I think that was copypasta too.

Fixed By rP20785

Fixed by rP20684

Fixed by rP20785

Fixed by rP23783

Fixed By rP20785

Fixed by rP23076 and rP23077