"Congress and policy makers can adopt as many requirements for standards as they like, but until such time that they recognize the challenges the industry is facing with our ability to certify current product upgrades, it makes no difference how many versions of new voting system standards are mandated if we are unable to bring those new products to the marketplace," Beirne said.

Brian Chess, chief scientist at software security vendor Fortify Software, disagreed, saying the U.S. still needs to make major changes. The U.S. needs "fully tested voting machines that reliably perform their functions, backup plans and well-trained poll workers," he said. "We are already hearing about unreliable machines, both DRE and optical scan, failing and causing long lines."

Chess called on Congress to pass "strict national standards on security" for e-voting machines. "We need to test ways the machines could fail and the reliability of the machines in a true election environment," he said. "We also need to write the standards to make the vendors responsible for the behavior of their machines, including the off-the-shelf components ... and when procedures break down."

Those standards need to be binding, he added. "No voting machine that fails to meet them should be used to cast a vote for our president."

The real question is whether fixing voting systems will be a top priority, even with "quite a few" reports of voting problems Tuesday, said longtime e-voting critic Eugene Spafford, chairman of the U.S. Public Policy Committee of the Association for Computing Machinery. "The question comes down to, how much are we willing to spend, and how confident do we want to be in the results?" said Spafford, a computer science professor at Purdue University.