The distance saving is the least useful number you have
Every Arctic route proposal opens with kilometres saved against Suez. Very few of them survive the questions that follow.
Great-circle distance between two ports is a fact of geometry. It is also, for a high-latitude route, close to irrelevant on its own — and it is almost always the first and loudest figure in any Arctic routing proposal.
The reason is straightforward. Distance converts to time only at an assumed speed, and speed in ice-affected water is not a vessel characteristic but a condition-dependent variable. A ship that makes 16 knots in open water may make 10 in loose drift ice and 4 in close pack behind an escort, if it moves at all. Multiply the same distance by those three speeds and the route arithmetic produces three different businesses.
Several other factors sit between the geometry and the schedule. Ice class governs what regime the vessel may enter at all, and the answer changes with the season. Draught restrictions apply in shallow straits along northern routes and can force a departure from the shortest line regardless of ice. Escort, where required, comes with waiting time that is not under the operator's control. Bunkering and repair options are thin: a routing that saves days but passes far from any yard has moved risk rather than removed it. Insurance and charter terms may price the route differently from the way the distance table does.
Operational perspective
A more useful way to state an Arctic routing case is as a distribution, not a number. What is the expected transit time, and what is the spread — the fifth and ninety-fifth percentile passage, given the ice regimes plausible in that window? A route that averages a week's saving but carries a two-week tail may be worse than the longer routing for anyone with a berth window, and better for a shipper who can absorb variance. The average alone cannot distinguish those cases.
The second question is what breaks first. Most Arctic schedules are not defeated by ice as such; they are defeated by a dependency that was assumed away — escort availability, a draught limit at the wrong state of tide, a bunker plan with no slack, an ice-class certificate that does not cover the regime encountered in the final fortnight of the window.
None of this argues against high-latitude routing. It argues for stating the case honestly: the saving is real, it is conditional, and the conditions are specifiable. A proposal that gives the distance without the conditions is not a plan, and the gap between the two tends to be discovered at the least convenient moment in the passage.