Wheels or slides?
I favour slide cars as the likely cart construction type, though I'd much rather it were wheels. My romantic notion about the temple builders is that they were technically very advanced and would have understood the advantages of wheels. On a long downward slope, a slide would be more easily controlled whereas over level or uphill stretches, wheels would be more efficient.
But the tracks speak for themselves. After all, their features reflect the nature of the carts that travelled on them. Hughes 1999 has written extensively on this topic.
Hypothesis 1: Slides The carts were of a type that required them to travel over a particular type of material, like sand or other slippery material, in order to slide smoothly and efficiently and with minimum wear. Moving these carts over rock was problematic, so cart ruts were carved in the rock to contain the slippery material and provide a much smoother passage.
Alternative Hypothesis 2: Wheels The carts were of a type that required them to travel over a smooth hard track. Moving these carts over rough rock was problematic, so cart ruts were carved out roughly and then filled and smoothed with cement-like material such as Torba, which has since washed out, leaving no apparent traces.
The ruts themselves may have been initially formed by grinding carts with slides made of abrasive hard rock. In normal use after that, the tracks were filled. The carts were pulled by oxen that walked in the filled tracks supplemented with human help in the steeper or troublesome sections.
4 wheels/slides or 2 wheels/slides? The answer to that can probably be found at San Pawl Tat Targa near Naxxar because that site has ruts travelling down steep hills and tight curves. A four wheel/slide cart will have trouble navigating the corners without at least one steering mechanism, but preferably two. When wheels or slides turn as a result of a steering mechanism, the distance between the tracks they make shrinks noticeably, so steered wheels would require tracks that narrow measurably around tight corners. If the tracks do not narrow, then the carts wheels or slides were not steered. This could only mean that it was a two wheeled (or 2 slide) cart. If they both narrow on the inside, and remain the same width on the outside, then it was a 4 wheel (or slide) cart with the front or back steered and the other end fixed. I do not know if anyone has studied the San Pawl Tat Targa site to make these measurements. Ventura and Tanti 1994 have made many detailed measurements of this site, but no published record of any variances of track widths around bends. I certainly intend to measure this site when I next visit Malta. Next to Clapham Junction, the Naxxar tracks are likely to yield the most information about how these tracks were actually used, by whom and for what purpose.