I've used a JBC very little (but did in fact try one) - but at home have a very old MX500 Metcal and love it.
Don't know about the watt ratings - since I can solder RF cages, ground planes, big lug capacitors, low gauge wire, even copper pennies without sticking - stuff a 40-50W iron should stick to rather than melt - but I have two thoughts about induction vs. direct:
1. Inherently faster loop: Curie effect is a physical phenomena of the material used in the tip. Hence the tip will always be at the Curie temp. Perhaps pumping 50W RMS into a physical load that "disappears" at the correct temp is substantially better than fast looping a similarly rated heater?
2. Heat transfer. One of the nice aspects of induction is that the tip is the heater. Almost as good as a soldering gun. There is no insulation, or a low conductivity (electricity & heat) material such as ceramics or Nichrome.
I used to use a Weller W60P that used a curie slug to control temp. It was excellent and I loved it. But it had a magnet that would turn the heater element on and off. The Power was there - but nowhere as instantaneous as a Metcal. The W60P had big insulators (it used a 220v heater) and was bulky and had heavy, bulky tips.
JBC made the tips smaller and increased the wattage so seemed to compensate for the magnetic control of the Wellers. But the telltale signs of insulating materials engineering are there:
https://www.eevblog.com/forum/reviews/normal-for-jbc-di-3000-to-overshoot-coming-out-of-sleep/ While you can make it smaller and smaller the latent heat of the Mica insulator is still there. So you overpower it with a aggressive wattage and aggressive proportional band - but something gives - you get overshoot (meaningless overshoot in the example above - less than 10% is nothing). Compared to a Weller - 0 lag. But compared to a Metcal - the lag is there.
If there is any "theoretical" lag in a Metcal - it would be when a part of the tip is above say 700F and the remaining part is under 700F. At which point the size of the heater becomes smaller. If you use a power meter (I built my own with an LM3914) - you'll see that the power supply just attenuates the output. Which is really cool since it doesn't use a thermal loop. It sense the energy used by the tip by measuring the VSWR of the RF signal. As I said - cool.
I did try to acquire a JBC - but couldn't justify the price of a new one (nor a new Metcal).
Availability of used, inexpensive JBC was sporadic at best (at least at the time I looked - and those that were available looked awful). For whatever reason people are dumping Metcal stations like there is no tomorrow (probably production lines and/or rework stations closing down). Mine was scratched with the name Gennady on it - but it is made of aluminum so a little cleaning and it is like new.....
Also tips - people sell large lots of used tips rather inexpensively. Sometimes even new tips can be had inexpensively. I did not see as many JBC tips on the used market.
I am not in any way disappointed in the decision I made but would have probably been just as happy with a JBC, a Pace or an Ersa (except for the part where I am able to say VSWR for a soldering iron - that is some major cool! Imagine a Smith chart for a soldering iron - stuff should be on America's Got Talent)....