Front loaders usually win on convenience and flexibility of placement options. Top loaders presumably have the potential of winning on durability and perhaps on average for nonresonance, though I don't know if anyone's actually got statistical or measured evidence for either assertion. Everybody's going to try to make the case that their solution is also the best sounding (or at least carries no penalty relative to the best of the other type). The reality is probably that too many other factors go into the sound quality and durability of players for broad generalizations about type to hold much water.
Anyone with more knowledge please advise me if I'm wrong here at all, but perhaps the larger divergence that's beginning to emerge in the disk playing field is between traditional mechnisms which acquire the data in real-time, and players (like the newer generation Meridians) using low-cost, high-speed computer DVD-ROM drives to 'read ahead' multiple times at accelerated speed and feed a digital buffer/interpolation/reclocking circuit which then feeds the converters in correct time, albeit slightly delayed. The approach is claimed to minimize the criticality of the load/drive/read mechanism in providing low-error, low-jitter data extraction, reducing the physical demands that it be built to some heroic standard with the attendant cost and replacement difficulty. (I personally have no experience to offer either way about this contention.)
Anyone with more knowledge please advise me if I'm wrong here at all, but perhaps the larger divergence that's beginning to emerge in the disk playing field is between traditional mechnisms which acquire the data in real-time, and players (like the newer generation Meridians) using low-cost, high-speed computer DVD-ROM drives to 'read ahead' multiple times at accelerated speed and feed a digital buffer/interpolation/reclocking circuit which then feeds the converters in correct time, albeit slightly delayed. The approach is claimed to minimize the criticality of the load/drive/read mechanism in providing low-error, low-jitter data extraction, reducing the physical demands that it be built to some heroic standard with the attendant cost and replacement difficulty. (I personally have no experience to offer either way about this contention.)