EZ81: That's bootyful! I think also of the great old Tek tube 'scopes, which had those lovely porcelain-solder terminal row affairs. They gave you a little captive roll of silver-bearing solder if you had to make the rare repair, beyond just changing out a tube.
Sand state changed our lives, given the higher transconductances and the importance of stray inductance, especially at very low impedances like at bipolar emitters.
At moderate impedances and frequencies I doubt there is too much effect arising from the lossy dielectric properties of glass-epoxy, phenolic, and paper-epoxy. However, years ago Tektronix, IIRC, making transistor scopes, did discover some very significant effects from glass-epoxy when they tried to make little caps from copper patches and ground planes. They dubbed the problem "hook" because of the step-response anomalies' appearance. It was very raw-board-batch dependent and went away more or less with the fine-tuning of the curing process. We have to trust that these processes are being maintained now if we have significant trace capacitances at moderate-to-high Z.
Air insulation is better than just about anything, unless you are dealing with way-high voltages. But unsupported stretches of conductors are subject to vibration and attendent microphonics in the presence of E and B fields, so keep things rigid, damped by something if possible, and short.
The prototype HDCD encoders and decoders had a lot of 3D air circuits in the analog circuits according to one account. I don't know what finally made it into production.
Teflon standoffs are good for really critical high-Z connection points, although even there we have a process issue: early batches had a piezoelectric effect that took many minutes to settle. We are talking electrometer-style currents and charges here though, at the ~100 femtoamp level if memory serves.
Brad