Taking some measurements, the scale is clearly wrong. Looking at the castle-shaped building in the section labelled 12 with a true-distance measuring tool (I'm using GIMP's Measure Tool), I find that its central block is about 12 pixels from river-side front to back. Measuring the scale using the same tool, 15 pixels is 20 feet, making the castle sans towers a mere 16 feet deep. That castle's footprint is smaller than my livingroom! And I doubt the intention was to model that castle after this one:
“A view of Broadway Tower” by Newton2, licensed under CC BY 2.5
All the other buildings have similar problems, with the smallest being 4′×4′. That's unbelievably tiny even by shack or shed standards, and I doubt they are supposed to be sheds anyway.
Clearly the scale is wrong, by an order of magnitude.
The trouble with many maps in big-name books is that there are many maps to produce, and typically these are handled by the art director as art rather than as true cartography. Drawing each bridge and building produces a particular map style that is labour-intensive and carries prestige, and is therefore sought after by art directors at big-name publishers. Rarely do the end readers actually try to orient on these maps in any but the most hand-wavey way anyway, and they do make the book extremely pretty. And, possibly more to the point, they make the book look how the buyer expects a campaign setting to look, with the text broken up by many maps.
Producing quality cartography is hard. So, quite probably there was no coherent intention of the kind you're trying to divine analytically from the map's properties, since any usability intentions were probably far behind business requirements like meeting production schedules and short-turnaround art orders, if usability was even a contender in that competition for business attention. Most likely, the maps are simply incoherent when looked at more than cursorily, and there is no intent — only the question of what you should do with them. And the easiest is to just use them abstractly, as a guide to layout and civic character.
The Feast of the Moon seems the closest. A summary from the Forgotten Realms wiki article:
The Feast of the Moon [...] was an annual festival occurring between the final night of Uktar and the first day of Nightal -- traditionally marking the onset of winter. It was also a time to celebrate and honor the ancestors and the respected dead. Folk blessed their ancestors' graves and performed the Ritual of Remembrance. People also gathered to tell stories of the deeds of their ancestors and of the gods until deep into the night, until they merged and become legend. This was a time to hear of past heroes,
great treasures, and lost cities.
In Faerûn, battles were typically fought between harvest-time and the coming of winter. This meant that most of the fighting usually occurred in the month of Uktar. The timing of the Feast of the Moon—after recently slain soldiers had joined the ranks of the dead—was thus practical, if sadly ironic.
Best Answer
Given that this is a lore question, and not a specifically 5e one, I would strongly recommend the 2e product Forgotten Realms Adventures, as it has a section on 24 cities of the Heartlands with exactly the kind of information you are asking for. The catch with it is that a lot of the listed characters there are now dead (if you are playing past the Second Sundering). But even then, in terms of the social structures, most of the data are still remarkably intact. For each city in the FRA, the following information is provided: Who rules, who really rules behind the scenes, population, major products, armed forces, notable wizards, notable churches, notable rogues' and thieves' guilds, equipment shops, adventurer's quarters (lodging), important characters, important features in town, local lore and finally a small map.
As for the 3e, FR Campaign Setting (FRCS) is excellent. At 320 pages and with a small font, it is one of the most information packed D&D books ever published. (It received the Origins Award for Best Role-Playing Game Supplement of 2001.) It provides a decent amount of information on the cities, but not in the itemized format of FRA. Instead it has data on regions/countries and cities are given as individual paragraphs. So it covers a larger geography (significantly larger than FRA, or 5e Sword Coast Adventurer's Guide), but provides somewhat less detailed data on particular cities. One excellent thing in FRCS3e is a map of trade between regions.
The 4e Forgotten Realms Campaign Guide (FRCG) is structured in the same way as FRCS. However it lacks the level of detail FRCS provides. The 4e "Points of Light" philosophy is arguably reflected onto FRCG and the particular data on the cities are very limited. Moreover, 5e D&D (and the in-game mechanics of the Second Sundering) has mostly returned the setting to pretty much how it was in 2e/3e.
Finally, the 5e sourcebook Sword Coast Adventurer's Guide (SCAG) has partially returned to the style of FRA (cities are discussed explicitly), with a mix of FRCS3e. It covers more limited geography than FRCS3e, and the data on the cities is less organized than FRA (and some data are missing). Its main pro is that it happens to be the most recent published material. If one did not care about editions, FRCS3e and FRA are more suited to answer the question you have asked.
PS: Both FRA and FRCS3e are available as watermarked pdf downloads that you can buy on the web for less than half the price of SCAG.