To all this good advice I will add that it's improper to use a PC board trace to carry significant current. Anything over perhaps 100 mA invites corrosion and trace separation. Making the trace wider doesn't help much, as the trouble is often at the ends of the trace where it connects to the outside world.
Interesting if true, however extraordinary claims require hard evidence, so please cite your sources.
What is the physics and/or chemistry behind the increased corrosion?
As a counter example may I point out the large number of consumer, industrial and other electronic devices with traces carrying currents in the 1A to 10A range? Provided they are designed to avoid excessive trace heating due to I
2R losses, reliability of the PCB itself is generally good to excellent. PCB damage resulting in failure is usually due to external causes, e.g. overheating due to restricted ventilation, surge induced electrical breakdown, or excessive mechanical stress.