Holmesian scholars have spent a great deal of the last century and a half pondering this question: Why was Doyle’s plotting so uneven? The fair-play issue set aside as mostly irrelevant – Doyle started writing long before the edict was issued – the truth remains that plotting in the Canon wildly varies from a story to another almost from the beginning. We are often told that the quality of the stories declined as Doyle’s interest in them declined too but it isn’t true, or not quite: gems, duds and indifferent stories can be found at both ends of the cycle. Doyle never slipped – but neither did he improve.
He didn’t need to. The Holmes fandom then as now was extremely easy to please; all they wanted was their Sherlock doing his Sherlock thing, no matter how well or how in keeping with the spirit of the original. Doyle knew what his audience wanted and he (reluctantly) gave them; he may have loathed Holmes but he also had to make a living. Bear in mind too that Doyle was not a professional writer – he didn’t see writing as a job but as a calling, and it must have been particularly infuriating and frustrating to him that he had to write « tripe » for money. Don’t expect him then to give it more polish than strictly necessary; don’t expect him either to care for things like internal coherence, hence the forever unsolved number of Watson’s elopements. Much of the Canon is Doyle in minimum service mode and by the end of his career he stopped pretending altogether.
Doyle as I said was not a professional writer in the way we now understand the term – someone able to deliver a well crafted product on demand. A Romantic by conviction and temperament, he relied on inspiration and it often served him well, be it either in well known (maybe too well known) stories like A Scandal in Bohemia, The Speckled Band or Silver Blaze or in less famous but equally rewarding ones such as The Naval Treaty, perhaps the most complex and modern plot in the Canon. Sometimes however his muse was tired or on leave and he rushed the job, reusing older plot ideas hoping no one would notice (no one did) or writing whatever he had in mind at the moment (too many examples, especially in later years)
All this I think accounts for Doyle’s erratic plotting, but there is at least one more explaining factor. Doyle was a mystery writer by accident. Unlike most of his followers he didn’t choose to write detective stories; he planned to write only one and then was sucked into the genre to no return. Having no particular ability or predisposition for it and few models to look up at, he was basically on his own and he made the best of it. This makes the sheer readability of most of the Canon all the more remarkable. It takes much more than skill or even talent to make something immortal out of what must have been the direst of chores. It takes genius – and geniuses can be forgiven for many things, including uneven plots.