Ill take a shot at this one, but someone please correct me if I am wrong...
Temperature is a relative value, and is normally thought of as being much, much more than it actually is. Yes, in order to melt metal, it does take a certain temperature, and an arc *can* output much greater than this temperature (could a spark plug arc melt metal?). The defining factor is electron density, or amperage vs. area. if you were to take a large plate or a small plate and push electrons through it with an even distribution (one big ground clamp and one big electrode), nothing would melt because the electrons are dispersed and are flowing easily. But when you require the electrons to jump across an arc with a small cross sectional area, there are a lot of electrons in a small space, reqiring everything in that space to raise in temperature (higher electron state) and melt. Perhaps raising the amperage will create more electron flow, but at the same time will increase cross sectional size, and give the exact same temperature. If you restrict the cross sectional area and then raise the flow, temperature increases.
Simply put, you cannot definitively say that amperage does or does not raise or lower temperatures. You can say that amperage plays a part as ONE of the variables that effects temperature. Others are the arc's effective cross sectional area, specific gas characteristics, heat transfer coefficients of the steel and atmosphere, characteristics of any solids being transferred across the arc, among many, many others. There just is no simple explanation. Think of a water jet cutting machine. The flow rate (amperage) of water is extremely low (how long would it take to fill a 5 gallon bucket with the water from the jet?), but the pressure and velocity is very high, and it can cut through steel. However, multiply that flow times one hundred (increase amperage) and divide speed and pressure by the same, and you get a waterhose, barely capable of ripping through a paper bag.
Hope that helps,
gls