Having missed the registration deadline for the GRE math subject test, I've been looking into the option of working for a year and then applying to math PhD programs once I have my test results. While I could try applying to the top programs I'm interested in this year, they generally say that my application will either be at a disadvantage or it will be disregarded entirely without a subject test score.
Since there are some (less competitive) US graduate programs that don't require the GRE, I was wondering if it would be a bad idea to attend one of them for a year and then try to transfer out? I thought it'd be a cost-effective way to take more math classes, but I read in this answer that:
It's difficult to transfer to a substantially more prestigious department. [...] No matter why you say you want to transfer, there will be some suspicion that your goal is to end up in a stronger department or a more desirable location. If you have another reason, you'll have to make a powerful argument for it.
So, I'm trying to figure out which idea is better: apply to a second-tier PhD program with the intent of transferring to a first-tier program, or find a job and just apply to schools after I take the subject test?
(Note: I recently asked a more general question about my options if I choose to take a gap year.)