[MrBungle]

Just to add, just because the design is targeting 5MHz doesn't mean it doesn't have several "accidental" oscillators running in the VHF region at the same time. The construction method can result in all sorts of weird behaviour. It's amazing when you first hook up an RF amplifier or something and wonder why the current is higher than expected resulting in much head scratching. Pop a scope on it and things are a little fuzzy. Out comes the GDO and "ooh it's oscillating at 7MHz AND 125MHz". That is if it hasn't exploded by then. This problem is quite bad when you think your average bog standard jellybean 2n3904 has a transition frequency of 300MHz. They can even oscillate quite happily at higher frequencies than the transition frequency annoyingly.

There can be a couple of pF between parallel conductors on stripboard, at 5MHz that's a reactance of ~16k which is probably not worth worrying about in a lot of cases compared to bias impedances. Out there at 125MHz, ~600 ohms which may result in some interesting feedback loops appearing, phase shifting and consequential oscillation. Padboard/veroboard is a similar story. All sorts of crazy gets out.

IMHO the best compromise on density and not being too ugly is using single sided copper upside down so the ground plane is on the top. You drill the holes through for the parts, soldering grounded leads to the top surface. If you don't want it grounded, you remove the copper around that hole with a larger drill bit, pass it through and do rat's nest on the bottom side. Good until you enter UHF when you need striplines and other black magic I avoid like the plague. I've had to use this for all the PLL and frequency synthesis experimentation I've been doing to keep things stable.