Well there is a difference between having similarity with other corona viruses and being identical. Take example of say human actin gene vs mouse actin gene. They are similar but not identical. There are enough differences in the sequence that I can design PCR primers such that I can specifically amplify the mouse gene and not human gene with that pair of primers (check out my qPCR post to have an insight into primers). Anyhow, same goes for viruses, there are similar sequences but you can always design primers specific to your virus of interest. Also, you have secondary test with primers for other genes such as RdRp and N. So even if you did end up getting a false positive due to presence of some virus which shares the sequence with your virus of interest (this issue is ruled out at primer designing step), the probability of this happening in both the cases is lower.
We don't directly go for sequencing because RT-PCR are specific they are robust, sensitive, and high throughput. You can literally carry out a thousand RT-PCR tests in parallel and get result in real time. Sequencing is more useful if you are interested in knowing the strains of the virus in the population than for detection. For instance if you have 1000 posts on hive to scan that are about Covid19, you can search if they contain that word/tag or not by using search or pressing CTRL+F. Why would you want to read 1000s of post manually to say - "ok, two of these are about Covid19". There it is you RT-PCR is ctrl+F applied to 1000s of samples at the same time.
Thanks a lot.
And I meant primer sequence, sorry to be misunderstandingly. Of course sequencing is way too expensive.