InfoStore seems to be what is generally called a 3-pane outliner (as Zoot). SQLNotes is a 1-pane outliner (with an optional 2nd pane) following the steps of defunct Ecco Pro. It can function as a 3-pane outliner when using the properties pane:
1- First pane is the properties pane, where fields are tree nodes
2- Double-click on a field will open items in that field in the ScratchPad, aka the 2nd pane
3- The HTML pane is then the content pane
Now the first pane, could also be simulated with a hierarchical grid list view. As Armando mentioned, this is planned.
-PPLandry
One thing that could probably be added is that there are really several ways of working with SQLNotes, depending on one's habits, aims, project, data at hand, etc. E.g. : Sometimes I use it as a 1 pane outliner (when working with numbers, bits of text data etc.), sometimes as a 2 pane outliner (when working with a lot of text and images), and I keep the properties pane (the third pane) opened... when needed, in all cases. And I often use several grids at once (very often it's the same grid shown in multiple "windows"), tiled vertically, for data comparison and analysis.
As to the concept itself, it is neither 1 table per grid (as this would restrict field assignments), nor a single table per file (as this would be terribly inefficient) and would not work well in a multi-user environment, which SQLNotes does very well.
-PPLandry
So there...

JohnFredC : What is it that you'd like to do with your data? Apart from the Grid tree (which Pierre said it could and would be implemented) and the "designation of Foreign Keys" (which can be simulated through field inheritance between items, set in the fields management window: "Value--> sub-items" — I use it all the time), are there other things?
PS: Note that there's a feature which will probably render grid hierarchical organization a bit less necessary than it seems right now: the coming "named filters". Users will be able to name their favorite filter combination and store them for easy retrieval, compounding, and use in all grids. What that means is that instead of having many different grids because of many different filter settings usage (or instead of rewriting filters or filter combination each time one needs new filters — even if there are already convenient ways to do that : alpha numeric filter toolbar + date toolbar), users will be able to store an abundance of different customized filters preset, more or less complex, easily recognizable and usable because of their custom "human" names.
These named filters will, IMO, function a bit as if there was a 2nd level of organization in grids... i.e. : sub-grids, so to speak. Or, in other words, different views of the data sharing the same Grid source.