How publishing works in this field
arXiv is the field's record. Posting at submission is near-universal in mathematics, establishes priority, and is how most readers encounter work years before an issue appears. Essentially all mathematics journals accept arXiv preprints.
Review is a verification task, not an assessment of impact. A referee checks a proof line by line, which is slow and cannot be rushed; a single round taking a long time is routine rather than a sign of trouble. Statistics and data science move faster and look more like a quantitative science, with code and data expectations attached.
What slows papers down here
Low citation density makes impact factors small across the whole field. A strong mathematics journal can carry a figure that would look weak in biomedicine, which is a property of citation behaviour and not of quality. Quartiles within the mathematics category are the only like-for-like comparison.
In statistics and data science the recurring objection is the opposite: method papers evaluated only on simulations the authors designed, with no real data and no comparison against the methods practitioners actually use.
Worth reading next
Questions we are asked in this field
Because the metric counts citations in a two-year window and mathematics cites slowly and sparsely. Compare within the mathematics category — the quartile — rather than against fields with faster citation cycles.
In mathematics it is standard, establishes priority and is how your work gets read. Virtually all mathematics journals accept it. In applied statistics and data science it is common too, with the same caveat to check an unfamiliar publisher.
Probably not, in mathematics. Verifying a proof is slow work done by one referee. A polite status enquiry after a long silence is reasonable; treating a long wait as a problem usually is not.