U.S. Air Force leaders in early 2021 all but admitted what critics have been saying for many years—the $100-million F-35 stealth fighter that Lockheed Martin LMT -0.4% is building for the service is too expensive and unreliable for hard, day-to-day use.

After 20 years of tortured development, the stealth fighter’s problems have gotten so bad that some officials have floated the idea of launching a new program to build a cheaper, more reliable fighter—one that finally might replace the Air Force’s roughly 1,000 old F-16s.

After all, it’s increasingly apparent the F-35 will never be as affordable and flyable as the stalwart F-16 is.

Just how big of a mess is the F-35? Dan Grazier, an analyst with the Washington, D.C.-based Project on Government Oversight, closely read the U.S. Defense Department’s latest weapons-testing report and came to sobering conclusion. The F-35 “remains in every official sense nothing more than a massively expensive prototype,” Grazier wrote.

He highlighted four problems in particular. In Grazier’s own words:

• Engineers can’t complete the Joint Simulation Environment facility. Taxpayers are paying a premium for the F-35 to be capable of defeating any adversary’s defense and anti-aircraft systems. The only way, short of war, to see if the F-35 can perform as promised is to simulate a modern threat environment. The contractor never delivered a functional simulation facility despite having had 14 years to do so, and the facility is still incomplete six years after the Navy was given the project.
• Program officials continue to struggle against a tide of F-35 design flaws. Nearly every time the engineers solve one problem, a new one is discovered. The F-35 still has 871 unresolved deficiencies, only two fewer than last year. Ten of these are the more serious Category I deficiencies that “may cause death, severe injury, or severe occupational illness; may cause loss or major damage to a weapon system; critically restricts the combat readiness capabilities of the using organization.”
• The F-35 program made some reliability improvements in 2020, but is still failing to live up to its maintenance and sortie requirements, despite the fact that those expectations were set very low. When aircraft are unable to fly often enough for adequate training, it can result in diminished pilot skills, increased peacetime accidents, and degraded combat effectiveness.
• For years, one of the biggest weaknesses of the F-35 program has been the deeply flawed maintenance and spare parts computer network called the Autonomic Logistics Information System, known as ALIS. Pentagon leaders finally admitted defeat in 2020 and pulled the plug on ALIS. It will be replaced with the cloud-based Operational Data Integrated Network, but the report warns that program officials are repeating many of the same mistakes made with ALIS, which would saddle the troops on the maintenance line with another flawed product.

So what went wrong? Lockheed and the Pentagon baked fundamental flaws into the F-35 program from the start in 2001, Grazier said. “It was based on technology requirements,” he said. “It wasn’t based on combat requirements.”

• US doesn't know what to do with Turkey’s undelivered F-35s
• The U.S. Air Force Just Admitted The F-35 Stealth Fighter Has Failed

In other words, the F-35’s designers started with unproven tech such as new sensors, fancy cockpit equipment and novel radar-absorbing materials and asked themselves, “What can we do with all this?”

The answer, it turns out, is a whole lot—but none of it well and never efficiently.

“It’s the exact opposite of how really classic and effective aircraft began,” Grazier added. In the case of successful designs, the developers “started with combat requirements in mind, based on actual combat experience—and the technology available at the time was used to make a very effective aircraft.”

Take the A-10, an affordable, survivable attack plane that remains as relevant today as when it first entered service in 1976. The requirements were clear—fly low and slow with a powerful gun and plenty of underwing munitions in order to support ground troops in even the most intensive close combat.

The requirements not only were goals—they were boundaries. To support infantry, the A-10 didn’t need to be supersonic like the F-35 is. It didn’t need to fly from an aircraft carrier like the F-35 can do. It certainly didn’t need to be stealthy like the F-35 purports to be.

In the case of the A-10, combat-driven requirements helped to produce a plane that did at least one thing very well and reliably—and at acceptable cost. That makes it pretty much the opposite of the F-35 with its ambivalent performance, poor reliability and high cost.

“The F-35 program is approaching a crossroads,” Grazier wrote. “The U.S. taxpayers are paying a premium for an aircraft that is supposed to be able to do it all. It is being tested to see if it can live up to the lavish promises made to sell the program. But that testing is revealing significant deficiencies.”

The Air Force already has paid for around 250 F-35s. It originally wanted more than 1,700 of the planes. It’s increasingly likely production will end well before that point.