Cranks but will not start?

The fact it runs on starting fluid proves the ignition system is working. For that to work, the automatic shutdown, (ASD) relay has to be getting turned on, so we know the crankshaft position sensor and the camshaft position sensor are okay. That leaves just the fuel system.

It is extremely rare for a Chrysler fuel pump to fail while you're driving, except for a wiring problem. Also, besides just hearing the click in front of the ASD and / or fuel pump relay(s), you need to listen for the hum of the fuel pump's motor. It should run for one second after turning the ignition switch to "run". If you can't hear that hum, listen right next to the tank while a helper turns the ignition switch on.

The way you described the feeling of running out of gas is exactly how this happened twice to my '88 Grand Caravan, both times at roughly 200,000 mile intervals. The elusive culprit was the pickup screen inside the gas tank. Don't bother with the external fuel filter along the right frame rail. Chrysler gas fuel filters typically last the life of the vehicle unless they rust out and start to leak. The pickup screen, or "sock" is a different story.

To make the diagnosis even more frustrating, the stalling occurs when the largest volume of gas is being pumped, which is during coasting. The engine will run much better at highway speed and when under load. The first time this happened to me, it took four hours to nurse it through Minneapolis with all three bypasses down to one lane for road construction. Once through the city, it ran great the next three hours to home until I slowed down for my turn off, then the stalling started again.

GM fuel pumps DO commonly fail in the manner you described, by running slower and slower. Given that failing Chrysler pumps almost always fail to start up, but once running, they keep running, I'm suspecting this sock, but before you tear into the tank, I recommend connecting a fuel pressure gauge to monitor what happens to fuel pressure when the problem occurs. I ran the hose under the back of the hood, then clipped the gauge under the right wiper arm so I could see it while driving. The second time this happened, I tied the gauge to the radio antenna for a year and a half. This also happened to my '78 LeBaron wagon and my '80 Volare, but those had carburetors, so the symptoms did include acting like running out of gas at higher speeds, but they would idle okay. It was at highway speed that enough volume couldn't get through. Those socks cost $3.00 each from the dealer. The sock for my '88 was $12.00 aftermarket, and just snapped onto the bottom of the fuel pump housing. I found that sock is not available separately for my '94 and '95 Grand Caravans, but those never caused this problem. I don't see just a replacement sock listed for your model either, so you may need to buy the complete housing with the pump and motor in it.

An alternative that I would perform for my own vehicle, but not for a customer, is to remove the pump assembly and cut a slit in the screen or just remove it. You'll still have the main filter on the right frame rail just ahead of the right rear tire, but there's one more detail to be aware of. Chrysler fuel pumps are built to very tight clearances to make them quiet, but that makes them easy to develop plugged impellers. It was common to go through four or five aftermarket replacement pumps, each failing within a few weeks, then, out of frustration, people would buy a new pump from the dealer and have no more problems.
The issue had nothing to do with the aftermarket pumps. In fact, NAPA for one buys their pumps from the same supplier that makes them for Chrysler, so you're getting the same pump. The problem was found to be microscopic debris collecting in the tank. It plugged the screen and what made it through was plugging the impellers. Each time the pump was replaced, some of that debris got collected and removed with it. By the time the dealer's pump was installed, all that debris was gone, so that pump didn't fail. The service bulletin instructed us to take the tank to a radiator repair shop to have it steam-cleaned when we had to replace the pump.

A friend with a small engine repair shop showed me a sample of today's gas with ethanol that had been sitting for a few weeks. It was cloudy and solids had settled in the bottom half of the sealed jar. He said that was mold that feeds on the ethanol, and that's what is plugging the screens. Remember, this was many years ago when gas formulations were quite different. I have ten-year-old gas in two of my cars, and they start and run fine. Today you're lucky if a can of gas lasts six months, so we know the additives are very different. They're much better in many ways, so that mold might not be an issue, but the screens can collapse too and block the tube where the gas is sucked in. The clue there is the engine stalls, and you coast to the side of the highway. After sitting in a pile of tears for as little as five minutes, the engine will restart and run okay for another couple of miles before it stalls again. (A failing cam or crank sensor can cause identical symptoms, but with those, you won't have spark or injector pulses, along with loss of fuel pressure, so the engine won't run on starting fluid).

You can find a fuel pressure gauge at most auto parts stores that rent or borrow tools.
For our friends with GM vehicles, especially trucks, those engines will sputter and stall or will not start if fuel pressure is as little as five pounds below specs. My '88 Grand Caravan called for around 45 to 50 psi, but the last time I finally got it to act up, I was dragging a tandem-axle enclosed trailer that is bigger and heavier than that van, and I could see the pressure gradually drop over about 20 to 25 seconds. That engine still ran fine down to 20 psi, which really surprised me. It wasn't until it hit 15 psi that I felt the surging and sputtering. I found that by lifting the accelerator pedal for just an instant, the pressure would pop back up to 50 psi, then I could push it again and go another 20 seconds. I nursed it 55 miles to home like that. Anyone following me would never have known I was having a problem.

I can't say all Chrysler engines will run down to 15 psi like mine did. Mine was a 3.0L Mitsubishi engine. Domestic engines might react differently. Also, for others researching this topic, older Chrysler engines with throttle bodies commonly run at 14 psi, so don't be surprised when you find that with the fuel pressure gauge.

Let me know if you think I'm on the right track, and what you find with my suggestions, then we'll figure out where to go next.