I noticed that the organization of fpDb is as follows:

list of wrappers (for each) list of tabIds (for each) object with internal data

Writing involves both wrapper name and tabId but reading is typically performed on tabId basis. Current structure requires that we iterate over all wrappers just to now if a hit was triggered in the tab.

It seems to me that a slight modification of the structure should provide benefits. I propose changing to:

list of tabIds (for each) list of wrappers (for each) object with internal data

This should be better for performance and also code readability.