Apologies for the delay in responding to Paralus’ well sourced comments but I think the passages are open to other interpretations.
Alexander’s actions after Granikos, and indeed the preliminaries to the battle (two scouting forces), do seem to have suffered from a less than satisfactory combination of his sources.
Let us consider the case of Kalas who is appointed Satrap of Hellespontine Phrygia at I xvii 1
καταστήσας δὲ Κάλαν σατραπεύειν ἧς Ἀρσίτης ἦρχε καὶ τοὺς φόρους τοὺς αὐτοὺς ἀποφέρειν τάξας, οὕσπερ Δαρείῳ ἔφερον, ὅσοι μὲν τῶν βαρβάρων κατιόντες ἐκ τῶν ὀρῶν ἐνεχείριζον σφᾶς, τούτους μὲν ἀπαλλάττεσθαι ἐπὶ τὰ αὑτῶν ἑκάστους ἐκέλευεν,
and is then sent to Memnon’s estates in the Troad from Sardis, with the Peloponnesians, except the Argives (garrisoning Sardis) and most of the other Greek allies.
[8] Κάλαν δὲ καὶ Ἀλέξανδρον τὸν Ἀερόπου ἐπὶ τὴν χώραν τὴν Μέμνονος ἐκπέμπει, ἄγοντας τούς τε Πελοποννησίους καὶ τῶν ἄλλων ξυμμάχων τοὺς πολλοὺς πλὴν Ἀργείων: οὗτοι δὲ ἐν Σάρδεσι κατελείφθησαν τὴν ἄκραν φυλάττειν.
Yet a half chapter later at xvii 1, Parmenion and Alkimachos are both given 2,500 ‘foreign foot’, who can only be these same Greeks, the Thracians and Agrianians appearing in Alexander’s force which moves against Miletos.
ἐν τούτῳ δὲ ἐκ Μαγνησίας τε καὶ Τράλλεων παρ᾽ αὐτὸν ἧκον ἐνδιδόντες τὰς πόλεις: καὶ ὃς πέμπει Παρμενίωνα, δοὺς αὐτῷ δισχιλίους καὶ πεντακοσίους πεζοὺς τῶν ξένων καὶ Μακεδόνας παραπλησίους, ἱππέας δὲ τῶν ἑταίρων ἐς διακοσίους. Ἀλκίμαχον δὲ τὸν Ἀγαθοκλέους ἐπὶ τὰς Αἰολίδας τε πόλεις ξὺν δυνάμει οὐκ ἐλάττονι ἐξέπεμψε καὶ ὅσαι Ἰωνικαὶ ὑπὸ τοῖς βαρβάροις ἔτι ἦσαν.
There are, naturally, more than one way of reconciling these accounts; I merely offer this as as likely an alternative as others.
I can see no reason for Kalas’ appointment to have been postponed until Sardis had fallen, with the battle won and Arsites fled the appointment surely belongs to the activity after the victory, the pardoning of Zeleia and ‘those barbarians who came down from the mountains and surrendered to him.’ These men are surely fugitive Persians from the battle, hence they are ordered to ‘disperse to their several abodes’. Alexander knows from these that the Persian army is no more and so can safely divide his command to send Parmenion to Daskyleion and detach the Greeks to ravage Memnon’s lands under Kalas. This will also have eased his supply situation, spreading the strain on the locals already close to supply exhaustion a week or so from the spring harvest.
Before Sardis had surrendered the detached troops will have rejoined him, and have been available for Parmenion’s and Alkimachos’ expeditions. Ionia was as yet untouched and the whole of Aeolis had not been traversed, even on the Coastal route; Smyrna, the satrapal capital lies south of Sardis. The argument that these expeditions speak of the territories not having been crossed by the army also falls by analogy with Asandros’ command in Lydia, which the army must have traversed if using the central Lydian route.
Ἄσανδρον δὲ τὸν Φιλώτα Λυδίας καὶ τῆς ἄλλης τῆς Σπιθριδάτου ἀρχῆς, δοὺς αὐτῷ ἱππέας τε καὶ ψιλοὺς ὅσοι ἱκανοὶ πρὸς τὰ παρόντα ἐδόκουν. I xvii 7
One could argue that the forces deployed, cavalry and light troops speak of a previously secured area, but I would say it points more to the nature of the terrain, mountainous.
None of these points is conclusive of itself, but I hope they demonstrate that the coastal route is by no means excluded by the ,admittedly peripheral, evidence of Arrian.
To Diodoros’ bald statement that Alexander went by Lydia two doubts may be introduced; we cannot be sure how the geographical unit of Lydia was understood by Diodoros or his source, it may have subsumed the language enclaves that Arrian calls Aeolis and Ionia; also, we can be sure that Diodoros’ source did not accompany the expedition and by analogy with the itinerary given from Susa to Ecbatana which seems detailed, complete with march lengths and halts, yet omits Opis and has the army cross the Euphrates twice, he had an inexact grasp of geography and the routes taken.
While I am defending Engels, I have to scotch Taphoi’s accusation above of optative emendment. The emendment ‘marganiam’ to Margianam is Vogel’s and perfectly fair, the line in question reads
Superatis deinde amnibus Ocho et Oxo ad urbem Margianam pervenit. VII x 15
Then, with the rivers Ochus and Oxus having been crossed he came to the city of Margiana (known) or Margania (unique).
Engels only notes p 104 n25 that at VII vi 10 the Mss read Marapunta for the editors Maracanda; no doubt you would accept the Mss reading and add another city to the gazeteer. Little wonder you have a well populated central route
When you think about, it free-choice is the only possible option.