I also subscribe to the "great filter" theory. About 25 years after the radio was invented, we were busy gassing each other in trenches, followed closely by a global pandemic, then mass genocide, then teetering on the edge of nuclear war. That's not a very wide window for aliens to notice our presence, if they rely on artificial radio waves to detect intelligent life.
My take is that technological improvements make radio sound like noise after a few decades. Early radios systems are very simple things which have signals (CW, AM, FM, ...) that are very distinct from electrical and thermal noise. Their signals were both drastically different from, and drastically stronger than, the background, enabling simple detectors to separate a signal's information from all that chaff.
Modern radios (such as spread spectrum systems, especially OFDM) squeeze nearly the Shannon Limit out of precious bandwidth (and also be frugal with transmit power) by using nearly all of it to carry information. This makes them virtually indistinguishable from a celestial object with a little extra heat (buried among things like stars, which have a LOT of heat).
It was only about 120 years from when Hertz and Tesla started making easily detectable radio waves to the Analog Television Shutdown, a significant milepost in the decommissioning of easily detectable radio signatures. I expect that, within anther few decades, the Earth will be emitting very little that might be recognizable as a radio signature of intelligent life, unless we expend a bunch of energy sending such a signature deliberately.
So my solution to the mystery expressed in the Drake Equation is that L (the length of time for which such civilizations release detectable signals into space) is short, not due to the falls of civilizations, but to economic incentives to use the aether only in ways that are no longer noticeable at a distance.