Look at the horizontal contour plot of my speaker measurements. The large majority of them are ±180° to about 300Hz - 500Hz depending on size, of course. But that's pretty much par for the course. (±6dB, normalized)
The KEF B2M is to about 300Hz. The KEF R11 Meta is about the same. The Neuman KH80 is to about 500Hz.
Now, if you want to get outside of the "norm" and talk about larger speakers with larger baffles then you can look at the JBL 4367 which is omni up to about 200Hz (again, ±6dB normalized).
Just to note, I was talking about ±90° previously due to the conversation being based on sidewall reflections. I want to make sure we're on the same page here.
There is a good chance that there is a terminology difference in our descriptions.
When viewed from basic acoustic principles and established terms an omnidirectional response is equal to all directions, in 4pi free space the DI will be 0dB when the response is truly omnidirectional. When the various curves are still diverging there is still some directivity and the response is not actually omnidirectional.
Where the darkest red colour in a normalized sonogram goes boom is where the box loses all directional control, just as there is the step above where the baffle loses control. These will vary with baffle width and box size and depth. The fact that the DI is not zero shows that the response is still not truly omnidirectional at that point.
You could call that point whatever you like, I'm not aware of any accepted term. What it technically is not is omnidirectional. The DI is often around 3dB at that point. There is certainly no significant directivity and to all intents and purposes the sound will be going everywhere. Semantics maybe but mixing up terms that have specific meanings creates confusion, and becomes the basis of accepted myths.
You may have seen in one of the Klippel graphics there is a visual showing where the different mode shapes become unreliable to describe the directivity. The monopole mode, n=0, starts to diverge at about 150Hz, in that graph. This coincides with where most closed boxes have a 0dB DI. Vented boxes are more complicated.