As to the relative performance of the two types of injection system...
Indirect injection means only that the fuel spray does not go directly into the combustion chamber, but rather into a pre-combustion chamber. There is still a high pressure spray of atomized fuel, but it is in a fairly narrow cone shape. This is because of the nozzle design, called a "pintle-type" nozzle.
With a direct injection engine, the tip of the injector actually extends slightly into the combustion chamber, and has multiple small holes (usually 4 or 5) for the fuel to spray through. It will be seen as a fine mist radially distributed in 4 or 5 sprays around the tip of the injector. This is called a multi-hole nozzle.
It's possible there is still a bit of air left in the hard pipe injection lines, which is why there is no visible spray into the mason jar. A small engine like this doesn't move a lot of fuel, even at start-up when it gets max volume delivery.
It may take 20 seconds of cranking time to actually get fuel out of the new injector until the system is fully bled. I would try loosening the new injector on the fuel line by a few turns, then crank the engine until you get fuel coming out where the line is loose. Stop cranking and immediately re-tighten the line and try again for at least 10-15 seconds. If nothing then, re-loosen the line and repeat the process.
It's possible you have injection pump problems, but the fact that it did this before and then started and ran fine after you fiddled with it, then re-occurred makes me wonder if it's not simply getting airlocked for some reason.
The injection pump has what are called "delivery valves" in the top of the pump which act as check valves to maintain a certain line pressure between injection cycles to reduce lag time when building enough pressure to open the injectors for each injection event. If the pump can make enough pressure to open these, and yet not enough to operate the injector, it's possible but not a common defect. In fact, I've never seen it happen before.
And for all the elements to be equally worn to this point at the same time is not likely either, ordinarily we'd see one or two damaged and not pumping, but not all of them.
I assume the glow plugs are working correctly?
Sean