Reading through this thread, I don't really understand why Occam's Razor is being thrown around so much, and this applies to both sides here. You people have to remember that Occam's Razor is fundamentally just an useful tool, not a rule, and it is often used when choosing between two hypotheses because simpler propositions tend to be far easier to empirically verify and test compared to more complex ones; it's not really the best thing to use when what exactly qualifies as "the simpler scenario" is extremely subjective and not at all a concrete matter, which is exactly the case here, and neither does it imply that the simplest hypothesis is automatically the correct one, either way.
Even outside of that, going out of your way to argue about author intent is completely fruitless and could be applied in both directions. Death of the Author is a thing, and if the author's intent isn't clearly expressed in the work itself, then trying to grasp at it for the sake of an argument holds as much worth as a random guess, which is to say, none at all.
As for my actual stance on it: As always, I think that the actual context behind an individual feat should determine its validity. If a character creates a parallel reality that is shown to be functionally just like our own, then going out of your way to say that a starry sky is just a background comprised of fancy lights seems like an odd assumption to me, especially if the reality in question, again, is an actual, physical place and not just some vague piece of scenery superimposed over a background, like
this, for instance.
This might sound a bit silly for some, but I believe the level of detail is another factor to take into account, though only when coupled with the above. If you can see a whole damn galaxy or a nebula in the skies of a dimension, then this is a different story from seeing a few white dots in the distance, and it'd take far more assumptions to say that something like this is just a product of fancy lights somewhere in the sky.