I don't think the odd nanosecond or two is going to be a source of trouble. If you've got traces that would impose a delay greater than 5 ns, I suspect you've got more serious problems than track length. Something's messed up in your routing/editing software.

Putting high-frquency signals on "fat" traces is probably not so clever, as it will increase the amount of radiated energy. Your neibors will "see" your computer's RFI on their TV. What you want is to keep as much as possible of the energy in your circuit on the board, rather than having it radiate in RF. Put your supply connections on wide traces, and, above all, around the perimeter of your board. Narrow signal traces and a solid ground plane on the side without traces will help a good bit. You can solder jumpers between vias on the solid ground-plane side of the board. You just have to relieve them from the ground plane. You may be surprised at how much difference that ground plane makes. Keep the supply connections between the large decoupling cap's and the IC's short, but don't be afraid to put the small cap's as close as 1 mm to the IC supply connections. Remember, too, that you don't need to remove any copper that could otherwise be used as ground or power distribution plane.

So ... you've got firmware being executed from your EEPROM? Does that mean your PC-based EEPROM programmer is fully functional? How, otherwise, would you have any display on your LCD?

RE