Neither/Both?
On the IGI if I put in Jean Edwards and Thomas Lyndsay as spouse I get 8 hits.
Of the 8 hits only one of them is an index from a parish record, the first one. The one listed as a marriage
APR 1827 Glenisla, Angus, Scotland with the batch number
M112904.
The other 7 entries are submitted by a member of the church, i.e. it is somebody elses research.
The positives are that there is a distant cousin researching the same family you could try to track them down through the church or through googling around to see if they appear on a Rootsweb archive or on Ancestry.
The negative is that you have no idea what they have based there information about Thomas's birth on. It may be guess work or it could be very good research - we just don't know.
With Scotlandspeople you are seeing the indexing of the original registers. There are loads of ways this can be mucked up to hinder your research. Bad transcription for the indexing or typos or just missed out?
When all your avenues are closed I'd use member submissions onto IGI as a guide to help you find the next record but do not use it as fact.
Your next step is to view the Glenisla marriage - it may give more information (or it may not).
Do you have census information for this couple? Track down every census - you may get lucky with other members of the family living with them, aunts, uncles, grandparents etc this can help you go back in time.
Do you think they lived after 1855 - if you do then find there deaths. If they died in Scotland then you will get their parents names, given that the person registering their death knew them.
Best of luck,
Pam