Guys really, wishful thinking, really. 0.91 sounds like an appropriate number for a 1.0 beta. We are a long way off that. Just some of the obvious things first (just off the top of my head),
Assembly WB in master,
better work arounds/solutions for topo naming,
I know it's OCC but fillets are not great,
also we get a lot of STEP imports that produce invalid solids, it happens a lot so although I admit that I don't know much about it, I am wondering how much of that is FreeCAD/OCC.
speed (and possibly file size) optimisations for large projects
better multi-document support
Possibly project file version control
Some defined minimum for CAM support, Path is great but I have no idea what minimum level of functionality it should have or if it is even there yet but still...
TechDraw, I assume it's due to toponaming, changes to the model tend to ruin the dimensions.
Better FreeCAD wide measuring system
Perhaps better add on version control in the add on manager, so for example you would get a warning when trying to add on a Python2 add on etc.
consider splitting add ons into "third party" and "officially supported" FreeCAD add ons, so that a user can have confidence that a add on will not go the way of Assembly2 etc. both in terms of "no longer working" and in terms of "compatibility", "certainty" of being able to read the files in years to come.
My personal suggestion is that if you really want to produce a real 1.0, then the first thing we need to do is change the way we work. We need to start attempting to drive the required development rather than the "whatever you feel like approach" that has admittedly served us well so far. I am not saying that we manage/demand like one could in a company/commercial environment, but we can "encourage" and at the very least "define" what we need/want, and hope that developers might look at that list and perhaps try to knock some of the item off the list.
If that is what we decide to do, then the first thing is to define what the specific "minimum" requirements are for 1.0, then advocate for the list to be worked on. I suppose before that we also have to decide what 1.0 actually means, to me it means a commercially usable feature "adequate" if not feature complete, application for a defined set of industries/use cases.
I also am a believer in that the features required for our "standard defined" industry/use cases, should be in master and not rely on vaguely supported third part add ons. But I would support segmenting FreeCAD such that you can install a minimal basic FreeCAD and then add the officially supported workbenches required for a specifically supported industry/work case, but without relying on 3rd party WBs for core functionality for those supported industries/work flows.
Anyway i have blabbered on enough for now.
Jim