The black hole information paradox from that article presumably fits. The conclusion from GR is that no information can escape, which is ultimately incompatible with QM, and that conclusion ultimately depends on the infinite density of the singularity.
I think the more charitable reading is that we'll find situations where either no sensible calculation can be done, or the sensible calculations we do churn out nonsense. Divergence in the UV catastrophe was an example of that, and Baez covered more here:
Struggles with the Continuum, https://arxiv.org/abs/1609.01421
1 Newtonian gravity
2 Quantum mechanics of charged particles
3 Classical electrodynamics of point particles
They explicitly mention they use point particles. Problem in 1 Newtonian gravity is especially egregious example. They model planets(!) as material points and see how they come infinitely close to each other. The author does mention these problems are pure math unrelated to reality.
Detecting (something about) the particle through gravity would at best remove the interference, not create an \infty somewhere in the model, implying that the model is self inconsistent.