I believe they're using "4 genes away" as a proxy of basepair distance, not similarity. It's all very messy business but it's not unreasonable that transcription promoters for those other genes also promote reelin transcription.
The place that I saw this in my own work was in the way that ends of intervening sequence (introns) was coded for excision during expression of proteins. The different ends of different introns behaved differently and introns in different parts of genomes had slightly different dialects.
The relationship of similarity and locality in regulators is different from this, but is another case of proximity being a proxy for similarity.