Charlie, a huge THANK YOU for taking the time to write such a thorough review. It's very much appreciated.

> In terms of multiscale/traditional, traditional is old and useless.  If this map feels like it will work, then it works

I was most concerned with the count-dependent rendering aspect, so you comment above is perfect and gives me confidence that this will work.

Great points on the other aspects too:

> The map title should be most prominent, and then the most important task widgets.

Good point, I'll see if we can tweak those settings.

> have a button for map use tips

Great idea. We wanted to avoid having a "user manual" but some tips and tricks would be useful.

> Draw the text on top of the thematic data

I'd love to be able to do this, but we need to use the Bing basemaps, which contain the labels, then we draw the polygons on top. Options could be reducing the transparency of the polygons, but then you lose the location context. (Some of the more traditional cartographers involved in the project didn't want any basemap at all - I think the current implementation is better than a blank screen)

> “fewer than 250”  and “more than 1,000” are not informative, and  " Your first legend class value should be 251, 501, 751"

I was applying a dose of the "traditional is old and useless" approach here. The difference between 250 and 251 isn't of paramount importance, in the context of an aggregated population statistic collected 5 years ago - the key point is that this suburb has more people than that suburb, and I wanted to avoid clutter on the legend. If people really care, the population pyramid allows them to see the exact values of total population and each population slice.

> The legend title is the field name?  Use Aliases

That's a bug - an alias should be read from the JSON. I'll look into it - thanks

> I would move the enumeration unit type below the legend elements. 

Good idea - thanks.

Thanks again for such a detailed review - stay tuned for the changes.

Cheers,

Steve