Since few seem to want to speculate on this, here are my thoughts on the matter. In the the lower latitudes during the day, the sun spends most of its time high in the sky so the light is coming from above most of the time. The best tree shape under these conditions is broad and relatively flat topped (think mimosa tree shape) with only
enough height to keep above the tree competition. Conversely at the high latitudes where the sun spends most of its time near the horizon so the light is coming from the side all of the time, the best tree shape is tall and narrow (columnar), which is the shape seen in trees of the taiga forests. In the mid-latitudes the optimum tree shape is intermediate between these two.
Our temperate trees weren't always growing where we find them growing today. From the time that angiosperms first evolved over 100 million years ago during the Cretacious until about 8 million years ago, temperate trees were growing up near the arctic circle as part of the polar deciduous forest (the conifer and ginkgo lineage in that location goes back even further). Then, starting 8 million years ago, as the climate cooled down into the Ice Ages, these trees were pushed south into the mid-latitudes. As they moved south, their shape evolved away from the columnar shape most efficient at the high latitudes into the broader shape that works best at their new home in the mid-latitudes. But those ancient genes for a columnar shape are still there and it is relatively easy to pick them out of the mix when selecting for new cultivars.
There are also narrower cultivars of tropical and subtropical trees such as avocado and Virgnia live oak. These aren't columnar, but are narrower than their typically broad wild type shapes and are are typical for the shape of mid-latitude trees, ancient genes going back to when these trees were growing in the mid-latitudes over 8 million years ago. But tropical trees have never grown in a high latitude location where a columnar shape would be optimum and so have never evolved the genetics for it, hence the dearth of columnar tropical trees.
Interesting, while researching this, I came across papers about fossil dendrochronology showing that the trees growing in the polar forests during the Cretacious were about twice as productive as their modern taiga equivalents, no doubt due to a combination of warmer temperatures, longer growing season, and higher
co2 levels at that time.
