It's a plausible idea, although admittedly it only solves one of the automobile-specific drawbacks and leaves the other issues unaddressed. Road trains show that coupling separate vehicles together yields efficiencies for the few, large prime movers, when compared to several small engines. Moving vehicles by rail is the most straightforward way to reduce friction to enable higher speeds.

But designing a road/rail pod that operates in two speed regimes would be difficult if neigh impossible. Consider crash-worthiness: a road-going pod doing 30 kph (18 mph) does not require substantial crumple zones (or at all) but the same pod on rails doing 120 kph (75 mph) will have to be at least tested to similar standards as passenger railcars. It's a strange beast to optimize for two very different conditions, like how airplanes are designed to be either sub-sonic or super-sonic, exclusively. Instead, something like the Amtrak Auto-Train which carries the pods at higher speeds would have all the benefits and nearly none of the downsides.

But going back to the benefits of this idea: platoons solve the issue of poor lane utilization due to spacing between vehicles fore and aft. Coupling is platooning with zero space in between, which also nearly solves the car-to-car communication issue, since they're now all physically connected. The same applies identically to rail, so the efficiency gains from lots of small pods to large platoons would be realized.

The problem that remain, though, are that these pods still need to traverse their destination: there must be pod roads, pod parking spaces and pod parking structures. Then pods will compete with active transportation and their lobby will seek to monopolize public spaces to the detriment of everyone that's not in a pod. Finally, zoning laws will enshrine the pod into ordinances requiring an obscene and arbitrary number of pod parking spaces by business type, inducing demand for pods when walking, bikes, buses, and trains would have also fit the bill.

All that has changed is that the pods can more efficiently flood the urban core and take up space. The 1960s American freeway building spree did exactly this with automobiles, and most cities have yet to recover.

It reduces the need for freeways without hurting mobility at the origin/ destination.

You are 100% correct, with the caveat that mobility is not hurt compared to automobiles. If the standard for public transport is to achieve automobile levels of mobility, we have already lost the game.