I trace my ancestry to a marriage in 1833 of Mary Dwyer to Martin Connors. I know their DOB, but rcds for the Cork/Limerick/Tipperary areas don't go back that far. Doubtful if we are related.
Co. Louth B rcds date from 1744 at the earliest and 1853 at the latest. Many parishes have gaps in their rcds.
For Thomas McCann B 1833 +/- 4 yrs in Louth, I found 3 B rcds: 1829 Cooley, 1833 Louth, and 1837 Tallanstown. But, none of them had a father named Robert. Therefore, it is unlikely that these are for your ancestor.
For Robert McCann, the earliest B rcd I found was for 1823 and 1808 for an M rcd.
For the surname Louth, there are 83 B rcds, but none for the forename Sarah. The same is true for marriages. You may want to explore the possibility that Robert and Sarah were born and married outside of Co. Louth: people did move about then.
I do find errors in the IFHF database: transcription errors where the yr 1827 is listed as 1287, listing of only one party to a marriage contract, typos, etc. My overall feel for the data is that it is better than 95% accurate. Remember, it can't be any better than the way it was originally rcd'd. Priests took a lot of liberties. I.E. Ths for Thomas, Pat'k for Patrick, etc. I try to take these known quirks into account when I do a search.