Background
To begin, I’ve been conlanging ‘properly’ for around two to two and a half years now and have been into it for far, far longer. However, I have yet to create a conlang that is truly in-depth, detailed, and realistic. I soon want to begin that process. I am confident in my knowledge of linguistics and have at least a basic knowledge of most areas and a more in-depth knowledge of others (none the less, I am still an amateur of linguistics, of course).
The current situation
At the moment, I am in the very early stages of planning – I’ve yet to decide on the final concept which I am going to develop. My ideas currently on the shortlist are:
A Celtic language spoken on a (fictional) island nation off the north-western coast of Ireland with North-Germanic influence due to interactions and mixing with the Norse people throughout history.
A Turkic language spoken in a (fictional) nation in the Southern Caucasus with influence from Georgen, especially on the phonology and vocabulary side.
A Uralic language spoken in Eastern Finland and the Karelia/Murmansk regions with Slavic influence.
Of these ideas, I am currently most drawn towards the first bullet-point, however, choosing the scenario of the conlang is not the crux of the issue. That is the issue of creating an in-depth and realistic, believable a posteriori conlang. I want to create a language that has detailed and believable phonology and grammar.
My questions to you
What is your advice on creating a good a posteriori conlang in general?
How do I effectively develop a conlang with as much detail and depth as possible?
Thank you very much for any help or guidance.