If you look at the flame, there's a hard central cone of pale blue flame, then a soft outer flame some way beyond it.
The hard central flame is where the primary air (the air that is drawn in at the back of the burner and mixed with the gas before it burns) is burning. The outer flame is where the secondary air (effectively ambient air) mixes with the partially-burnt gases from the primary stage and continues to burn.
The overall flame temperature is affected by the Air:Fuel ratio. Hottest occurs when all the fuel burns with all the Oxygen from the air. In forge terms, this gives a neutral atmosphere.
Going leaner (more air), reduces the flame temperature (the extra air absorbs some of the heat produced) and gives an Oxidizing atmosphere.
Going Richer (less air/extra gas) also reduces the flame temperature (the extra gas absorbs some of the heat produced) and gives a Reducing atmosphere.
If you have a rich flame in a forge, the secondary flame, where the hot gases burn with ambient air, happens outside the forge as dragons breath.
You have very little control over the primary air with a torch, but you can control the secondary air to some extent, by altering the position of the torch to vary the amount of air getting in around the outside of the torch.
When I made my 2BF a few years ago, I made the burner hole conical (Though TBH, I don't know how much this helped, as I didn't try any other arrangement). With the torch pushed in to the narrow end of the hole, there was very little secondary air gap. backed out to the big end, or even just beyod, there was a big secondary air gap.
With the combination of mixture adjustment and gas pressure adjustment, I was able to get forging temperature with no problem and could (just about) get down to a usable O1 soak temperature (805-810 degC; 1480-1490 degF). I was using the biggest head of 3 from a cheap no-name torch set. I had lots of dragons breath at the HT temperature and less, but still quite a bit, at forging temperature.
Interestingly, I was able to get up to welding temperature by changing just the jet in the torch for the one from the smallest torch head. Much less dragons breath.
Type 23 Insulating Fire Brick melts, by the way. I built my next mini forge with 1400 degC CF blanket.
I used a Mineral-Insulated thermocouple with a handheld readout to make the adjustments but took it out to actually use the forge.
I later found I could get much better control with a purpose-built burner incorporating a sensitive choke. This provides all the adjustment needed on the primary air and gives only two variables; gas pressure and the choke setting. With the thermocouple to provide feedback on what is going on, it's relatively easy to get things to where you want them. My burner used a commercially-available adjustable Venturi, which I'll concede is probably cheating.
Later still, I used pretty much the same burner in a drum forge made from about 18" of 10" pipe lined with 1" CF blanket (I used 1" CF board for the ends because I had it laying around, but blanket and sheetmetal would have done the job just as well). The idea was to make a scaled down version of a Don Fogg-style 55 gallon drum HT forge. It seemed to work ok. I could get adequately consistent 01 HT temperatures for as long as I would ever want them and over the full area of any blade I might ever make.
With the rich burn and reducing atmosphere, scaling doesn't seem to be a problem at all. If you can get a good burner, this would be my recommendation for a cheap HT setup in a small shop. Use it outside though; rich burn means Carbon Monoxide.
Bear in mind the temperatures in the pics are in degreesC
http://i667.photobucket.com/albums/vv31/timmgunn1962/P2240429_zpsbd62d989.jpg
http://i667.photobucket.com/albums/vv31/timmgunn1962/P2240428_zpsc728e337.jpg
http://i667.photobucket.com/albums/vv31/timmgunn1962/P2240427_zps4fbcc6e8.jpg