why is the non-inverting op amp an obvious choice for the design of a high impedance voltmeter?

Answer 1

This belongs in engineering, not physics.

It's not an obvious choice, but it is a good one. It's a complete, high-gain amplifier in a single component. Inversion wouldn't matter. You would want one which is DC coupled so you can avoid the complication of adding a chopper circuit to measure DC voltages.

Answer 2

Recall that op amps are always used with negative feedback. The only exception I can think of is when you want to use an op amp as comparator. In this case, positive feedback is used to advantage.

Otherwise, negative feedback is always present. This means there's a resistor placed between output and inverting input. This feedback resistor makes difficult to achieve high impedance at this terminal, even when the feedback resistor itself may be in the MΩ range.

Op amp circuits are generally designed for gains over 100 and higher. Necessarily, then, the "input" resistor (going from inverting input to ground) has to be that much smaller in comparison. Since the highest commercially available value is about 10 MΩ, this means that input resistor is to be in the range around 100 kΩ. If higher gain is required, input resistance should be even lower.

I recall that, when discrete transistor amplifiers were used, a feedback technique (bootstrapping) was often used when high amplifier input impedance was required. But since op amps have two inputs, both potentially usable, it's much more simple to opt for using the non-inverting input as such.