There is absolutely no way you will achieve a 50% increase in torque (at any RPM) just by switching from ordinary port injection to direct injection. Do you have any references to this claim?
The efficiency improvements are possible, mainly due to the better atomization of the fuel, but 15% seems pretty generous if we're talking about improvements from DI alone. Keep in mind that many manufacturers are adopting multiple new technologies at the same time they're switching to DI. Newer engines with DI also tend to come with wideband o2 sensors and more modern ECUs, both of which can lead to improvements in efficiency which makes it harder to determine the true contribution of DI.
There's only one possible explanation for the 50% torque improvement claim:
Many cars are switching from larger naturally aspirated port-injected engines to smaller turbocharged direct-injected engines. Maybe that's what they were talking about.
Going from a naturally aspirated port-injected 3.8 liter V6 that makes 300HP and 280 TQ to a turbocharged direct-injected 1.8 liter 4 cylinder that makes 300HP and 350 TQ could very well explain those figures, but it's the turbocharger and smaller engine that are really the key here.
They can control and modulate the injection far more with DI, because they're not limited to intake injection. I cannot give you much in the way of credible sources on the 50% claim from DI, only that I've read it from multiple sources - so treat it as marketing speak.
But even if we're "only" gaining 5% efficiency from DI, it's a pretty big deal. I'm fairly sure a GP3 or GP2 engine supplier have stated that DI provides better mileage, power, and torque, especially with an FI engine, compared to MPI.
The efficiency improvements are possible, mainly due to the better atomization of the fuel, but 15% seems pretty generous if we're talking about improvements from DI alone. Keep in mind that many manufacturers are adopting multiple new technologies at the same time they're switching to DI. Newer engines with DI also tend to come with wideband o2 sensors and more modern ECUs, both of which can lead to improvements in efficiency which makes it harder to determine the true contribution of DI.
There's only one possible explanation for the 50% torque improvement claim:
Many cars are switching from larger naturally aspirated port-injected engines to smaller turbocharged direct-injected engines. Maybe that's what they were talking about.
Going from a naturally aspirated port-injected 3.8 liter V6 that makes 300HP and 280 TQ to a turbocharged direct-injected 1.8 liter 4 cylinder that makes 300HP and 350 TQ could very well explain those figures, but it's the turbocharger and smaller engine that are really the key here.