I don't buy this. Local snowcover definitely decreases surface temperatures, especially during clear, windless nights. So weather models, especially back in the early 90s, might have over-estimated temperatures when there was snowcover. It might also shift marginal wintry situations slightly colder, although this is questionable. But I believe that any significant impact of regional snowcover on synoptic scale track/features is anecdotal and not scientifically well supported.