With WASAPI you are getting the level of the source.
If you are using Spotify’s volume normalization* many/most tracks will be quieter than with normalization disabled. That’s because you can reduce loud tracks, but many quiet-sounding tracks can’t be boosted without clipping (distorting).
Your levels are not THAT bad. You can use Audacity’s Amplify effect if you want to boost the volume after recording.
I think Spotify calls volume matching “normalization”, but technically normalization usually means something else. (Normalization sets the peak to some preset level but peaks don’t correlate well with “loudness”. Most digital music is peak-normalized, including quiet-sounding songs.)