I do have a suggestion for filled zones. I stumbled over this problem a few days ago.
Let me start with an example that is trivial and would not justify this request. But it makes the case clearer.
Often enough, you have zones, that do represent completely isolated regions on the board. For example on an SMPS. One side is mains, the other side is power output. Or a µC controlling mains over opto-couplers.
This example is quite obvious. We need isolation between N and L1 (left side of the layout). So the isolation is set to 0.5mm (all on layers 2 and 3). Then there are 3 opto-couplers bridging the gap between mains and the controlls and an AC-DC-converter. It is a no-brainer to make that isolation gap as big as possible, at least using the OCs as guide. But this also already makes clear, that the isolation gap on top and buttom layer need to be bigger than on sandwitched layers.
Now to a more useful example:
It is an isolated USB-interface (IIRC 1 kV). The voltage difference within the highlighted area is 5V. So clearance within the zone is 0.2 mm and enough. But the clearance to any neighboring zones needs to be way bigger. And it isn’t in this example. I simply forgot about it.

So my suggestion is, that zones should have an extra parameter “boundary isolation”. By default, it is chained to “isolation”, but can be unchained and set to say 2 mm without affecting isolation within that zone.
By “chained” I mean something like this:
You can click on the chain, and then assign a different value to the boundary isolation. A check box would work too.
There are other examples, where isolation is more embedded within a PCB. For me, that is often RS232 or RS485.
I hope, I made my point clear. If not, feel free to ask!
So, is that a stupid idea, am I just to lazy to adjust the gap manually or …?