An important point to note is the following: To make their mass & radius measurement, Doroshenko et al have to make the assumption that the neutron star is emitting uniformly in all directions.
If the neutron star had only one or two bright spots, that would lead them to underestimate its size (given they assume uniform emission).
They justify this assumption with the fact that they do not see any pulsation from the rotating neutron star.
BUT that could also be explained if our line of sight is aligned with the stars rotation axis. That way the angle under which we see any hot spot stays constant and so does its apparent brightness.
They estimate they probability to miss pulsations by chance for such a non-uniform atmosphere and end up with 0.8$-$21.3$\,$%, depending on its rotation period and atmospheric model (see table below).
The 21.3$\,$% corresponds to very short rotation periods that might be unlikely.
For the most typical rotation periods of a few hundred milliseconds they get 4.36$-$0.81$\,$% , with 0.90$\,$% being the probability for the more realistic model with smooth temperature profile.

So they say:
We conclude, therefore, that the absence of observable pulsations due to unfavourable orientation is also unlikely in this
scenario, even if cannot be completely ruled out for a single object
for arbitrary temperature distributions.
What they mean is, that they also have not seen pulsation from most other similar sources and you would not expect to always get a lucky alignment. So looked at together, these sources probably do have a uniform atmosphere and emission profile, even if you cannot be certain just from this single observation. For this they also reference Wu et al 2021, who conclude with:
We estimate it is unlikely ($< 10^{−6}$) to attribute that we do not see pulsations to an unfavorable viewing geometry for five considered sources.
So, is this a "real and confirmed observation"? Hmm... probably. But I would say their uncertainties on mass & radius might be a bit underestimated as they only account for statistical uncertainties, after assuming perfectly uniform emission, which is not guarantied.