For Thomas Gannon B 1838 +/- 1 yr, somewhere in Ireland, with the father forenamed Thomas, I found 1 rcds: 1837 Tipperary.
For John Gannon B 1840 ditto, I found 1 rcd: 1839 Sligo.
For William Gannon B 1843 ditto, I found 1 rcd: 1842 Dublin.
There is no discernable pattern here. These results are all likely false positives. Either the B rcds for these people didn't survive, or haven't been transcribed yet. Suggest you wait until the database is more complete.
Do the English M rcds have any info as to placenames and/or parents' names?