The Himalayas and the extremely mountainous and jungle-filled terrain in modern-day Myanmar is the top culprit for the North vs South Asia question. To this day land travel through those regions is extremely limited.

The second is that the Mongol Empire and successor states did a tremendous job of consolidating and supporting the Silk Road into how we think about it today. They built Caravanserais at resting points in and in-between the oasis-towns to support trade in addition to protecting it. (and of course taxing it)

The second is that sea travel during the heyday of the silk road lacked many of the comparative advantages that it gained in later years. Between storms, pirates, and the need to sail all of the way around Indochina, the risks were pretty tremendous for little or no gain. That said, those sea routes did exist, especially for trade between the Indian subcontinent and the West. and of course, those sea routes eventually did replace the silk road with the arrival of european traders.

The Song Dynasty in China often found themselves in a position where they had to 'pay off' horse nomads who inhabited their northern frontier* to sustain peace in their empire. Part of this pay off consisted of large quantities of silk. In that way much silk found its way to a northern market from where it could be slowly traded west. There was thus an impetus that encouraged the existence of a northern route.

*: It's very likely that such Sino-Barbarian transactions occurred at other times too, but I am not knowledgeable enough to speak on it.

This is actually true, and a there was an enormous maritime trade in the Indian Ocean basin in late antiquity, the Middle Ages, and in the early modern period. However, until very recently this maritime trade was basically ignored for several reasons. First, there are simple practicalities of documentation; the corpus written evidence coming out of India, Southeast Asia, Arabia,and Africa is smaller and less-studied(in particular, there are fewer Western scholars with the training to study that corpus) and historical archaeology in India, the horn of Africa and the Swahili coast, and in Arabia is a very young field and not as much has been done. Even with the limited archaeology that has been done, however, valuable data has emerged-we've found for instance Chinese pottery fragments on African tombs, substantial archaeological evidence of contact between Ethiopia and Arabia(for instance, inscriptions of Ethiopian kings that use the Greek, Ge'ez and Old South Arabian scripts), a mysterious but intriguing Indian statue found at Pompeii, and other archaeological evidences of extensive trade networks. This evidence supplements existing written records of trade in the Indian Ocean basin, such as the various cargos recounted as coming to the Red Sea via India in the Periplus of the Erythraean Sea and the travel accounts in the Christian Geography of Cosmas Indicopleustes. In addition, many archaeologists and historians historically assumed that there was nothing of historical interest in sub-saharan Africa, which was one of the major centers of the late antique and medieval Indian Ocean trade, and archaeologists of the Near East tended to be less interested in both the archaeology of southern Arabia-which was assumed to be "tribal" and not worthy of study-and in excavating non-monumental architecture. Needless to say, the increased archaeological research at Swahili sites like Kilwa Kiswani and Songo Mnara has done a great deal to disprove this belief.