As a quick example, I recall Earthborn used the fact that Objectivism assumed a persistent, knowable reality as part of the argument against Objectivism being a practical philosophy.
Sort of. Here's what I actually wrote: "If a philosophy has "central tenets" that are provably untrue (such as "human beings have direct contact with reality through sense perception, that one can attain objective knowledge from perception through the process of concept formation and inductive logic"), then it fails almost immediately." I expected someone to respond to it a bit sooner.
It is provably untrue that human beings have direct contact with reality through sense perception. Every optical illusion shows this: our contact with reality is rather indirect, filtered and distorted by our senses, coloured by our biases and earlier perceptions. No amount of "concept formation" or "inductive logic" can change this into "objective knowledge".
However I'm fairly certain Rand acknowledges that this is an assumption, but argues that it's a necessary one.
Other philosophers before her (Kuhn, Popper, etc) have argued quite convincingly that it isn't a necessary assumption. Objective reality does not need to be truly knowable even to make scientific progress. Instead of trying to prove what ideas about reality are true, one can also make progress by showing how some ideas are more difficult to prove untrue than others even if they are falsifiable.
I would argue that the assumption is even a dangerous one. If one assumes that reality is knowable, and it turns out that it isn't, then one is further away from reality then when one assumes it isn't knowable and it turns out that it is. The former assumption made one believe in a lie, the latter makes one humbly admit that one didn't know everything.
If reality isn't knowable then there's nothing for us to base any of our decisions on, and it doesn't matter what we do.
If reality isn't knowable, it can still matter a lot what we do. Maybe not to the universe as a whole, maybe not to God, but it may still matter to ourselves.
I don't think humans have access to "objective reality", but just because there is no ultimate truth, does not mean every untruth is equal. Some untruths are untruer than others. Just look at the Asimov quote in Roboramma's sig.
We will never know the true objective shape of the Earth. What shape we imagine it to be is even dependent on subjective preferences, such as how we choose to measure it, or how precise we want our calculations to be. It still matters what we do, and just because we can't prove some answers be right, we can still say with some confidence that some answers are wrong. The Earth is for example not a cube (no one ever says the Earth has 8 corners, people from all 4 corners of the Earth agree it is a tetrahedron.)