Okay, I'll provide a formal answer, Genbank phylogeny is to get a gist of the diversity. It isn't a formal phylogenetic analysis. You can down the tree as a pdf, I don't think you can download the treefile (in fact I'm pretty certain you can't). Genbank will perform a reasonable distance method and from recollection it will "collapse clades", so if you see little triangles on the tree it represents a large number of taxa represented as a triangle.
I was surprised to learn there are 4000 SARS-CoV-2 sequences.
To make a tree you need to download the data, align (muscle) and send through a formal phylogeny program. If you think Genbank is okay, then try MegaX its very point and clicky. Generally, maximum likelihood phylogeny has returned to vogue for at least 1 decade now (previously Bayesian). However, the Beast package is a Bayesian analysis which is still going strong and the main tool now used in molecular dating.
In summary, to perform a tree analysis beyond Genbank
- use the download option (there are two variants - the entire genome or the cut locus based on the query sequence)
- align using muscle or clustalo (muscle is really quick(
- perform a tree analysis using MEGA X, phylml, raxml or Beast depending on what you objectives are