Two things about this strategy:
1. Barbarians early in BNW are very strong. If you are unlucky you may have to go for archers very early which will slow down your progress.
2. Food. You may restart enough times to get a good first city that can eventually go 30+, however nothing can ensure you that you will get optimal locations for the other two cities. And with no military you cannot contest good locations near enemy civ's borders.
So, this is a good strat, but it is highly dependant on map and luck. I tried it with Korea, I got my first two cities (I had to clear a barb camp to settle the second city, so I had to produce an Archer).
My map however does not have a good spot for a third city near the first two. So either I have to settle a crap city, or delay and settle too far away.
Bad form to quote a whole post, for which I apologise. However, Red_Doctor's experiences mirror my own very closely.
Barbs are strong and a PITA. I have had to change the build order to include an extra unit prior to the second settler. Which type obviously depends on whether or not I pop Archery from a ruin.
Getting 3 good sites is another issue. However, I think I prefer to play with 4 or 5 cities (personal preference). Even though city 3 is poor it can grow and contribute units. The next two cities can be captured. I am able to ignore AI second and third cities and just capture capital cities. This is because I can produce a flood of archers, spears and cats really fast. So 4 of 5 cities are extremely strong.
I am practicing this start quite hard. I find it quite robust, and you must adapt to the circumstances. This is not painting by numbers.
This is my latest Empero start - As Korea by coincidence.
Points to note:
1. At turn 107 I have only met one Civ.
2. I have built cargo ships in lieu of caravans
3. My third city is quite food poor.
I was unable to get enough cash to buy worker + granary. Therefore I bought the granary only.
Barbs were a real nuisance and I put and extra unit in the build queue. Together, with this and the slight delay by building cargos instead of caravans I was able to save more money.
The explosion in growth meant that at turn 55 I had three cities, and turn 62 three cities with libraries, and turn 71 I had NC.
After Philo I was researching ancient age techs at 4 turns per.
After NC I was researching classical techs at 7 turns per.
By turn 90 I had a trireme, 2 cats, 6 archers and 3 spears.
I had 600g in the bank to upgrade all the archers. In the end I couldn't do this because I am a complete newb and didn't know you couldn't upgrade units outside your own territory. I bought a colloseum instead.
Turn 100 I had captured vienna + the worker I couldn't buy.
It is now turn 107 and next turn I will get a circus and a colloseum to cure the happiness issue.
The upshot of all this, is that despite being a complete newb, and despite a very average start, I think I'm in great shape. In particular my rate of research is just ridiculous compared to any of my previous games.
for little 'ol me to succeed then this must be a killer strat IMHO. There may be better ones around, but I haven't found one yet (and am definitely not clever enough to invent one).
A couple of screenies.
on one of those. 
to start growth with an average of 1 national wonder found per game at this stage.
But, eh, how do you manage to get the money for both granary and worker and bribe? That's 700+ in total, it's kind of heavy money for the early game. How much did you bribe for, BTW?
