What you lose in national wonders going wide you gain by simply already having the slack in production. The measly 8 production from ironworks is easily made up for by a city sized around 6 population ideally; when liberty has 3 more cities each sized population 15, this production is massively compensated for. Sure, each city increases tech costs marginally, but you'll be by far making more beakers for each city to compensate for the slack lost. National college usually has a beaker value altogether of about 10 or 15 total when it's built; its hammer value can be as much as basically two caravans with a reasonable number of cities, and it takes only a total of 10 population with libraries to compensate for the NC's beakers. NC I prioritize roughly before turn 100, but that isn't too hard with 8 or so cities. If you build your last expand by turn 75, work the right tiles, you can get the last library by turn 85-90; national college in the capital shouldn't take more than 8 turns, unless you have a no-hill no-pasture start. University-rushing tradition finds itself losing out in BPT to wide play with competent internal trade management and worker strategies.
Realistically, liberty players find their happiness and gold dipping twice in the early game: once very early, this is after you spam your first several expands, and it can be fixed with the right luxury focusing and cs appeasing, and by working plantations and gold camps with spare population. Of course, even after this, once you've got your steady cities ring and they all have their pop up, there comes another dip right around the late medieval, right before finishing liberty. However, this is around when you're completing markets and getting the circus maximus up. I'd say the build order I usually pursue in expands goes monument granary into either colosseum, market, or library, depending on where I'm weak.
With the second to last policy in liberty you either pop your free golden age early or grab the happiness from city connections, depending on which problem is more severe with the first dip; usually it's the city connections. This means finishing out liberty, you'll either have a massive empire finally getting enough happiness to sustain a massive growth, or just entering into a golden age during which you get your roads up and use production boost to spam colosseums and markets. Either way, the admitted problems caused by early expansion are fixed by the late medieval and you're ready to focus the potential you've built up.
As to playability and balance, it's great because wide grabs up higher base yields and those base yields are then compounded massively with the right percentage buildings, resulting in a much larger raw output than tall, providing a very tangible and immediate advantage in faith, production, and gold (which do not scale) and roughly being even in science, culture, and happiness (which do scale). Come ideologies it edges out and ends up overcoming every single comparable facet of output and is simply superior to tall over time.
As to historicity and realism, I find it very realistic! Modern superpowers are all huge states with power derived from sheer output, the U.S., China, Russia. Sure, they lost out on certain advantages to smaller core empires in earlier time periods, but come industrialization size was crowned and the infrastructure fostered following expansion became compounded massively with new ways to focus raw output. Over time, wide beats tall.