1. We will find alien civilizations by their radio signals; and
2. The limiting factor on alien civilizations are Earthlike planets.
(1) is the flaw in SETI. The general argument against this goes something like this:
1. Planets are efficient ways to store matter (in that gravity binds it together) but are highly inefficient in creating living area or collecting a star's energy;
2. Entering and leaving a gravity well is expensive (in energy terms);
3. Gravity can be trivially replicated with centrifugal force using materials we already have (eg stainless steel).
4. Within 1000 years we will be capable of building space habitats that solve the above 3 problems;
5. As a consequence of the above, the natural tendency for any growing civilization will be to encompass a star with orbiting habitats.
6. Artificial structures orbiting a star, even if they're just energy collectors, would be incredibly obvious to any observer from a huge distance and would be far easier to detect than radio transmissions similar to what we do today. The tl'dr of this is that the only way to get rid of heat in space is to radiate it away. The signature of this is a function of the temperature of the object and for a huge range of temperatures this is in the IR spectrum and this huge IR signature without a corresponding visible light output is what makes it "obvious".
The first radio transmission occurred 120 years ago. In 1000 years (and possibly well before them) we'll be beyond them in terms of the signature alien civilizations would detect (in that, they'd see our IR signature first). This is of course the blink of a cosmological eye so what we're really doing is hoping to see civilizations who are at the same point we are, which is incredibly unlikely.
This ties into (2). I even saw Neal degrasse Tyson make this mistake in a talk where he posits the answer to the Fermi Paradox is that civilizations run out of planets and die out.
Planets are likely important as a cradle of life and the number of suitable planets is likely a filter of some sort but irrelevant in the long term. A star with no planetary system at all can easily be a home to a Kardashev-2 civilization. Matter can be extracted from stars directly (as a side note, extracting Helium to avoid the star going supernova is likely a useful side benefit) and/or constructed with particle accelerators (and all that energy).
What I find compelling about this argument is that it's /almost/ just an engineering problem (albeit a huge one) rather than one requiring huge advances in technology and/or "new physics". Steel is sufficient to build O'Neil cylinders and solar power is a sufficient energy source for all this (ie it's not predicated on the viability of commercial fusion power generation, which I for one am not yet convinced of, and if it is viable everything gets a whole lot easier).
The other attractive part of this is how it can be done.in a piecemeal fashion, meaning you can build one orbital, the another, and another and another. Certain other megastructures are much more "all or nothing" (eg ringworlds).
The above is the classic Dyson Swarm, originally called a Dyson Sphere but Swarm tends to be the common name now as a bunch of people have assumed (incorrectly) that a Dyson Sphere is a rigid sphere. Dyson never suggested that and no known or even theorized material could support a rigid sphere that large.
People tend to hone in to one issue or another with the above. A popular one is about waste heat. "What if they recycle it?" So with perfect recycling they've violated thermodynamics so we're just talking about increasing efficiency, at which point you've just reduced your IR signature, not eliminated it.
The natural consequence of all this is that we are relatively alone and likely the only technological species at or above our technology level within our cone of effect in the Milky Way.