We need some semiconductor industry expert, anyone here?
Asked a friend at Intel. TLDR you can't really compare it directly
Conventional techniques don't use lasers, so it's already not a 1:1 comparison
Fabs also measure equipment speed in WPH, Wafers per Hour, not nanogap electrodes per hour, since complete litho machines work on a defined wafer size (200mm or 300mm wafers). Given that this is a theoretical/experimental technique, there's no actual implementation right now.
Furthermore, nanogap electrodes are....esoteric, to say the least. So far everything I have read indicate that it isn't a full transistor, but rather used to join molecular components together or used as part of a transistor.
Assuming that each nanogap electrode represents a single transistor, this would mean that it's at a rate of 50k transistors per hour. Way too slow to be useful.
In comparison, EUV machines at Intel outputs around 100 wafers per hour, with each wafer having more than a trillion transistors, which makes me question whether the electrodes are being used as part of a transistor or not, or why they decided to report the speed using that particular measurement.