So, this won’t be the most scientifically accurate description of what’s going on, and my own understanding and word usage may be imperfect, but hopefully it gives you at least some idea of what’s happening.

With your low-pass filter you’re making your signal reject frequencies any higher than the cutoff point that you have set. All it does is eliminate those higher frequencies without modifying anything about the frequencies lower than, or at your cutoff point.

When you begin to add resonance, what you’re doing is accentuating the frequency that your cutoff point is set to. So not only are you rejecting the frequencies higher than your cutoff point, you are making the frequency at the cutoff more prominent in your remaining signal. The frequencies below the cutoff point will remain unaltered other than at a lower resonance value you will see a bit of emphasis at frequencies immediately below the cutoff frequency as the signal is drawing those frequencies up sympathetically with the boosted resonance frequency. As you increase the resonance you are further accentuating the frequency at the cutoff point to where the signal will show more of a spike at the cutoff frequency rather than the slight rise seen at lower resonance values. This narrows the band of any sympathetic frequencies that may have been previously affected at lower resonance values. While the lower frequencies are still present and unaltered in your signal since they are below the cutoff, the prominence of the frequency at the cutoff point is being amplified so drastically in comparison that your signal is dominated by the resonated frequency.

I hope this makes sense. I found it difficult to know if I was saying things in as clear a way as needed.