Thanks for the note!
Economies of scale lead to increasing returns indeed, but at some point in the pre-digital economy they start to wane under the law of diminishing returns: too many employees (numerous employees are better at demanding pay raises: ask the American car industry) + scarcity of essential inputs (lithium for Tesla for instance) + overstretched infrastructure. That is the problem when you’re stuck on the Northern Side.
In the digital economy, traditional economies of scale also wane at some point, all other things being equal. But since network effects + machine learning + the architecture of participation continue to drive increasing returns (the magic of the Southern Side), a digital company can find that second wind and continue to grow anyway, thus sustaining some economies of scale beyond that threshold that no traditional company could ever cross (Facebook’s 1.5B users for instance).
All in all, the expected increasing returns are much higher in the digital economy, which explains why any new entrant is willing to drop prices even lower than what traditional companies did in the 80s—because if it wins (and it’s a big “if”, since there usually is only one winner), it will ultimately operate at a scale so large, from a position that will be so dominant, that it will figure out some way to make even more money and finally recoup those early, unreasonable price drops.