Many railway modellers do not have direct knowledge of the design and practices of the prototype railway. Instead, most of us have to rely upon advice and information gleaned from manufacturers, 'experts' or the Internet. In my case, the use of points, the lack of traps and the spacing of sidings was based upon a professionally designed track plan for a previous model, many track planning publications and even manufacturer's information sheets.
The rationale behind my track plan is the track adjacent to the platform is the main line. The next one is the passing loop. I accept passenger trains cannot cross but we are talking of a small, rural station in 1910 where passenger trains would be few and far between (they could be timetabled to pass at an adjacent station). The third line is for a coal train to drop off wagons and collect empties. The locos originally envisaged were a shunting loco, a fully lined Dean Goods and a fully lined Small Prairie. And before anyone says anything, all GWR locos were fully lined up until 1906 so a fully lined goods loco in 1910 is a reasonable supposition. Looking into something else unrelated, I discovered that shunting operations in small, rural goods yards were carried out by either manpower or horsepower right up until the 1950's (the last shunting horse was retired in C1967 at Lambourn). The shunting loco idea was scrapped and the model adjusted to include stabling for Hercules the Horse.
Regarding scenic blunders, I was thinking more of kits or ready made houses where the position of the chimney would indicate the fireplace was behind the front door, yard offices and the like where the door cill is flush with the ground when the reality would be a minimum of 6" up stand, etc., etc. If we are discussing overall modelling faux pas, I have seen videos of much vaunted layouts that include impossibly steep gradients to both tracks and embankments and many other such errors.
But I return to my original point: whether the model railway is correct to the prototype is secondary to the satisfaction of the modeller. We model for our own enjoyment not to produce 'museum' quality masterpieces correct in every detail although we try to follow the prototype as best we can. From this perspective, any criticism of someone's model requires it to be constructive rather than destructive; encouraging rather than it being a 'put down'.