73 valiant runs good then randomly dies

On a 73,the very first thing to do is disable the seat interlock switch.If everything then works, you are done.
But if not;there is a simple thing called a one-spark test. It has to be performed immediately at the time of failure.
It goes like this;
A) Do not turn the key off when it fails. Pop the hood, pull the coil wire out of the cap and near-ground it, about 1/4 inch to the engine (any unpainted part). You watch the gap while a helper turns the key off.Pay attention. You should observe exactly one spark, when the key-switch breaks the circuit. You can cycle the key as many times as you like, and there should always be just the one spark. If you have the spark as described, then theoretically there is nothing wrong with the primary ignition system. The following parts are all working; the coil, the ECU,the ignition switch and everything between it and the battery,the ballast resistor, and all the connections. Pretty cool eh? Only the pick-up is still in question, and the secondary system.That's the one-spark test, and it becomes the tool for further testing.
B) To test the pick-up; remove the distributor cap. You will have to draw an iron object across the pole piece of the magnetic pick up. every time the iron passes thru the field, a spark should occur. The easiest way to do this is to position a reluctor vane at the leading edge of the pole piece, by manually turning the crankshaft until the alignment occurs.Then with the key on and the coilwire still near-grounded, you rotate,(using the rotor), the a reluctor-vane past the pole piece. A spark should occur every time the vane passes by the pole, in either direction.
These tests should only take a few minutes.
You should practice them, when you know the engine will run, to familiarize yourself with the procedure.
C) Now that brings us to the secondary system, which is everything else from the cap to the plugs, and those you have replaced,already.
D) But, what if at A above, it fails the one-spark test. Well the first thing usually is to look for a voltage loss. So that means you have to carry a VOM or test light in the car. But even easier is to clean the damyn grounds. The ECU will not work if the case is not grounded back to the battery. That means all points between the ECU and battery must be cleaned. I think there are 3 or perhaps 4 points. Then repeat the one-spark test. That Ecu could be leaking it's sealer goo all over the apron and that's no never mind; it can continue to fire the coil for thousands of miles and many years,pay it no attention.The pick-up does not have a ground and no attempt should be made to install one, and the D itself does not need to be grounded. In fact, you can pull it right out of the engine, spin the driveshaft and watch a stream of sparks issue from the near-grounded coilwire.Same for the coil, the case needs no ground.
E) but say it still won't fire.And you have replaced a HOT ecu, indicating by logic that it is seeing voltage. Turn the key off.The next thing I would do, (if no VOM is available) is to cool the coil off, and retest the one spark deal. This is usually done by flooding it with water. A little messy but water evaporates. This will take more than a glass of water splashed on it. The electrical part inside is divorced from the metal case, and a cooling oil bath carries heat between them, the case being the heat-sink. If your coil is mounted upside down,and it appears to be leaking it's oil away, just replace it, it's gonna cause trouble,lol. But say it's physically OK, then we have to cool the electrical part inside. So you trickle water on the case. The cooling case absorbs the heat from the oil, allowing the hot electrical parts to transfer more heat into the oil. Eventually all the heat will transfer to the water flowing over the case. But you don't have to wait that long.Just trickle a quart or two of water over it and back to the one-spark test.If the spark returns badaboom!, crank it and look for a stream of sparks. If you get it, yore work of diagnoses is done.The coil appears to be bad. Wait...appears to be? wel you don't know if you cooled the coil, or if it had previously cooled to just below the failure point,just before your test.Right? So now you have a choice; just replace the coil. Or,get that coil hot again,until the engine dies, and this time go straight to the cooling water test.
F) but if cooling the coil makes no difference, then you are gonna have to prove the coil is getting power.I always do this first (as you did with the power-jumper) cuz VOMs are getting to be so cheap now, I have several and there is always one in my classic cars. But until this moment, it hasn't been an absolute requirement in the diagnostic. So,hook it up and, if you get a little under battery voltage with the key on and a cold ballast yer good to go. Or are you? No you have to wait until the ballast gets hot and retest. Leave the VOM in circuit and watch the voltage drop as the resistor heats up, which might take three minutes or more. The voltage should drop continuously and drop down to somewhere around 8 volts. Close enough. Now turn the key to "start"and the voltage should pop back up to near battery voltage. If you get a pass, then remove the factory wire from the Minus side, turn the key to "run" and check the voltage there. It should be a hair less than the plus side, maybe a volt less. The point is it should NOT be Zero. If it is, throw that coil away. Replace it and back to the one-spark test.
G) But say the coil passes all the tests and the voltages are good. Now you have 2 more choices. Is it the ECU or is it the pick-up. Back in B) above we did a quick test of the pick up.Now we are gonna do an electrical test on it, cuz pick-ups are cheaper than ECUs. The pick up can be tested using the AC setting on your VOM, You are looking for less than 2 volts AC ouput during manipulation of the magnetic field at the pole piece.The pick-up is an electrical signal generator, depending on the disruption of the magnetic field by the iron reluctor or any iron object passing thru the field, to generate it's little signal.You can do this with the same set up as in B) by just disconnecting the connector and probing the pick-up side.
This test works best with an analog-type VOM, because you can see the voltage rising and falling as the iron passes over the pole. If you are using a screwdriver,it will slam onto the magnet as you wave it over, don't panic, just continue the stroking, it makes no difference. It works best when the vane is within a specified range of the polepiece. The usual range is .008 to .011. I have examined running cars, and I have seen gaps from zero to .030. So this gap is perhaps not as critical as some mechanics will say, but I'm sure Mopar has a reason for the spec, and it's easy to reset. If using a digital VOM, the digital numbers will flash by in rapid succession from zero to peak and back, so if you see something close to 1 volt, and flashing numbers, you at least know it's doing something. And if it is, that just leaves the ECU
H) you can't test an ECU except by process of elimination and substitution. Which is why I used to always carry a spare. But in ten years I never saw a failure, so eventually I threw the spare back in the box with the others.
But you can test the wiring to and from it. Now you can do that two ways. You can test the appropriate terminals in the connector for the correct range of voltages to it with a HOT ballast, And you can test them for continuity between the ends. If the connector proves good, you only have the ECU left to blame.
On your car, you can temporarily sub in ANY other ECU, from any other car, and back to the one-spark test. If it passes that, then the cranking test. And if it passes that. then you can order up a new one, or just leave the tester in there,lol.
Parting shots;
J) You gotta remember, this is not rocket-science. There are exactly 5 parts to this test; the coil, the pick-up, the ballast, the ECU, and all the wiring,including the ECU ground. Only the ECU is non-testable. And finally, as someone else already mentioned, I think, the cure on these old cars is all too-often in the ground or the wiring. But I have to give you credit with your power jumper. You pretty much proved a problem between the input side of the HOT ballast and the coil, or the coil itself,lol.
K) BTW, The dual ballast has ,well, dual resistors.One steps the voltage down to the coil in run. The other is a power reducer to the ECU. So one resistor can fail, and the dang thing can still get hot,fooling you into thinking it's working.The block gets its power source from the same blue wire, and you can see the little jumper right there. So if one side gets power, logic says the other will too,lol. But logic ignores the stoopid spade connectors, and 45 years of corrosion and atmospheric attack on the copper. So this is one of the biggest sources of troubles..............
besides that stoopid seat-belt interlock device.

I hope I got it all.