That's right, but unless you connect MV behavior with human factor. Simply MV's fell "victim" of their own reliability: Most lamp on the end of their life or die completely and/or start to behave annoyingly (burner explosion, cycling, color changes, do not strike,...), what ensure responsible people take care to relamp. Failure to do so yield or to nonworking fixture, what (if properly designed) does not consume power, or strong user complaints (when cycling). In contrast MV's only gradually glow dimmer, so slowly, it is not noticed by anybody, while still drawing full power. The point, where the lamp output fall below some threshold (in recent years industry chooses 70% as an economic brake-trough), is then defined as lamp's end of life, so when lamp should be replaced. The fact the loss of brightness is so slow mean, the user should set and follow relamping schedule. On the other side the high percentage of working lamps at this "age" mean, such relamping might be limited to ONLY scheduled slots, e.g. in the scope of trough site maintenance. But because people are people (and not machines), such replacement of "still working" bulbs is often considered as a luxury, so an operating cost, what might be postponed. The result is then low efficacy lamp (with double the rated lifetime or often more burnt age), what really consumes power and does not provide much light...
What is really irony here, today's "green flagship", LED, has the same habit: Never totally burn out, but gradually, trough it's life, loose efficacy. So my guess, they would be accused of low efficacy in the exactly the same way as MV's, unless their ballast would incorporate something like "End of life low efficacy cutout" shutting them completely down...
|