I don't really have an answer to your question, but I can give you some starting points. First regarding whether you are in the soma the answer is "very unlikely". Keep in mind your typical acute recording electrode will be ~100um in diameter, at least ~50um and up to ~250um for linear probes. Chronic recording electrodes (e.g. Utah arrays) can be much smaller but never smaller than ~10um. The body size of neurons vary within the range 10-50um. So if you poke the soma of a neuron with your electrode you will pretty much kill it instantly (unless you use patch-clamping, e.g. https://www.nature.com/articles/nmeth.1993/).
Now dendritic electrical potentials are small (few mV). Only after being non-linearly summed up will they reach the threshold that will trigger a spike (https://www.annualreviews.org/doi/pdf/10.1146/annurev.neuro.28.061604.135703). Spikes is what you are able to record, it is unlikely that dendritic potentials will be high enough for you to detect above background noise. Even if, it means your electrode is close enough to the soma and your recording will most likely be dominated by the neuron spikes, which will have a much higher Signal-to-Noise Ratio.
The trickiest part of your question is whether you are recording from an actual neuron or its axon. I see no reason why you wouldn't be able to record from an axon. Statistically I do not know how likely it is, considering that axons are far smaller, but much longer than the soma. I can provide 2 thoughts: (1) does it matter? Axons within gray matter most likely come from local neurons. As a physiologist your goal is to know when/what/how the neurons within the vicinity of your electrode are spiking. So does it matter that you are recording from the soma itself or its axon few hundreds of um away? (2) When you lower you electrode too much and reach white matter you start seeing strange spikes, usually I see "double-spikes" that look like an "M". Based on the depth of the electrode, background noise and that this doesn't look a regular spike, it is quite clear that this waveform comes from an axon. I've personally never seen similar waveform within gray matter. But of course I cannot be 100% sure that I've never recorded from an axon either.
As for how to keep your sanity, as always in science you draw conclusions based on what is most likely. Physiology is very hard, so as long as you made the best possible job targeting the correct recording area, isolated good-looking waveforms, did a rigorous spike-sorting and found meaningful effects related to your stimulus/task. Then is it more likely that you recorded mostly from proper neurons, with possibly few artifacts here and there, or that your recorded mostly from dendrites or axons from far-away neurons (that is things that are not related to computations actually being carried within the area your are recording from), and that all your effects are just an artifact? You'll never be able to completely eliminate uncertainty, but you can reduce it to levels that make you quite confident what you are reporting is true. Asking this question is a good start though.