There seems to be some discussion as to what is 'proper design' - as to what is 'proper' depends on the standards or design criteria you expect to comply with. Many of us need to comply with CE but we then have industrial vs commercial requirements. My stuff is industrial so the radiation requirements are less but the susceptance requirements are higher. My equipment comfortably passes the radiated tests to industrial spec (Class A??) but would fail the commercial spec. To pass this, I would most certainly need a 4 layer board. For fast transients, my boards regularly exceed the test. So CE is really the baseline as you need to pass it for legal reasons, but if you know your equipment is being placed into a situation where interference exceeds the CE requirements, it is wise to design your equipment to meet these. also, you might have application specific standards such as aerospace, military or automotive to comply with. In Farshad's situation, even a PLC might have problems - if the installation isn't up to spec, then the best designed piece of equipment may not survive. it's a bit like having the world's best piston connected to a standard con-rod in an F1 engine. It's no use blaming the piston when the broken conrod smashes it through the crankcase.

As my previous post, I think the solution would require attention to a number of items, not just Farshad's board. Although I would think there are some improvements that could be made there that might not entail a four layer board. Again, it depends on what the compliance requirements are.