The answer to this is both yes and no.
The reason I say yes and no is that SARS-CoV-2 exists as what is known as a quasispecies. Quasispecies are the population of virions that exist in an infection, with each separate virion containing a genome with its own mutations as a result of natural error rate of the RNA dependent RNA polymerase. When we sequence a virus and call the result its genome, what we are actually looking at is a consensus sequence, not the individual sequence of each virion that went into the sample from which the genome was derived.
Now, the problem with a quasispecies is that not all the virions produced in an infection are actually viable and can go on to produce more virions. In some cases there will be defective genomes, some more will have defective capsids or defective packaging of the genome and so on. These together make up a series of defective particles with a range of different terms depending on how they behave in an infectious process.
In many cases there are defective interfering particles, which miss parts of the genome or have genomes that are otherwise defective but are capable of infecting cells, but not capable of producing infectious particles themselves, and can compete with functional particles for infection (hence the interfering part).
In some of these described particles and within the sphere of the quasispecies will be virions that contain new ORFs. How those ORFs behave in an infection is very difficult to say as they are very hard to study. In most cases these new ORFs will cause a failure of some other (already) critical function and produce non-viable virions, but without examining any potential new ORF functionally in a lab we can't be certain that a potential new ORF does anything or indeed what the product might do.
For an estimate of the coding capacity of SARS-CoV-2, you can look at this study (1), published in Nature, which revealed that there probably are unknown ORFs within the virus. However, I suspect that these were there already and are not newly evolved as such, just newly identified.
- Finkel et al., Nature 589, pages 125–130 (2021)