I would really like to see what article you have that shows diesel being even remotely close to electricity for efficiency?
I'm sure if you change the objective to 'economic efficiency' instead of a simple physics or engineering efficiency, diesel power becomes much more 'efficient' and closer or even overtakes electricity.
Electric Trains vs. Diesel Trains
Though trains are more efficient than trucks, not all trains are equally efficient. Diesel-powered trains transfer about 30-35 percent of the energy generated by combustion to the wheels, while supplying electricity directly from an overhead powerline transfers about 95 percent of the energy to the wheels. Powering trains with electricity rather than diesel has several other benefits ...
The statement above is ... IMHO ... very misleading as it conflates electrical transmission efficiency with electrical power generation and distribution efficiency.
All electrical power generation has a thermal efficiency. Generally, in larger scale the efficiency improves, but the actual thermal efficiency varies greatly depending on the technology of generation used. The idea behind electrification is one big hunking steam turbine is more thermally
efficient than 100 little steam engines aboard locomotives.
But ... there are losses distributing the power. These vary greatly depending on the location of the power stations, the size of the grid, the weather, what else is on the grid etc. *Then* there is the 1500V vs 25kV. The later should be more energy efficient, but it may not be. I'd like to see some hard data on that.
There are also capacity factors to consider. Most of the power generated by a locomotive is actually used at source for it's intended purpose: moving a train. The total potential power output of a big hunking power station is not always used, because it can't readily change it's output to match demand. Typically it's still around 90%.
@RRT_RULES might like to help us out here, there are tables on the all the transmission loss factors for each generator on the NEM, as there are for the distributors - all published somewhere
But essentially, the argument for diesel is the "small" generator just the right size for the task at had is not as thermally efficient as the large stationary generator. No question about that. But it's not more efficient by a factor x2 - though perhaps an older diesel with a 30% efficiency vs a CGST generator of ~60% is the extreme case. In some cases (Victorian brown coal ?) the off site generator can be less thermally efficient than a modern diesel.
Then, there are the losses from capacity factors, transmission and distribution.
These numbers are just fabrications to illustrate the point, but if:
Large scale power generation is 45% thermally efficient.
Diesel generation at 35%.
But there are 10% losses due to capacity factors, 10% in transmission and 10% in distribution, you are better off simply generating the power at source.
Electric traction is way more efficient that direct steam, but I don't think the case for diesel is as clear cut.