This thread raises questions that make for interesting discussion but in my opinion the answers for same can’t be given as generalized firm conclusions. This is a “boxers or briefs” underwear question – and like all such questions the best answer in my view is as Bob Dole famously resolved: “Depends”.
I would grant as a premise to the discussion that your system can’t sound any better than your speakers – just as it can’t sound any better than the program material you choose to play.
Speakers are that fundamentally important. If you don’t pick superlative music to play, the output of the system, in the end, can’t be superlative, and by the same token if your output transducers are not superlative… well you get the idea.
I would further agree that program reading components [CD players], amps, and cabling can all be had in the current market for very low costs on a comparative basis versus the cost of pursuing excellence in speakers. This is not surprising given that all the gear upstream of the speakers has become by comparison trivial to produce to exacting tolerances and tight specs thanks to modern manufacturing techniques and the state of the art in electronic circuit design. Speaker builders at the top end of the food chain on the other hand don’t have equal access to the electronic wizardry and mass production advantages that for example a CD player builder does.
Because of the fact a speaker by definition must interact with the real world [move air in a controlled fashion] one simply doesn’t have the ability to short-cut around big problems inherent in the laws of physics with mass produced IC chips or a fat read ahead buffer.
If you want a speaker to produce tight controlled low frequency sound you have to use a high quality low frequency driver [not a trivial thing to produce], and you have to mount it in a very stiff box built to exacting standards… design a x-over circuit that matches the abilities of the LF driver, precisely mount HF elements, taking into account time coherence… and on and on... all the while implementing these design elements at a real world scale. In short, you can’t micro-process your way around the challenges of sound production at the speaker end of the system with anything close to the facility one can use such techniques further up stream.
As a result I would consider it silly for person to divide up their system budget on anything approaching an equal basis between electronics and speakers. The rational choice to me is obvious… you heavily weight your budget to getting the best speakers you can afford.
But when it comes to acoustic treatment of a space, one runs into a very similar set of choices. Your system can’t sound better than the room you put it in, and in small rooms [less than 7,000 cubic feet or so – on the order of 19x37x10’] the laws of physics leave you with conditions that electronic wizardry cannot address. For example, the modal density in small room results in the inability of an untreated space to produce smooth, tight-sounding low frequency resonances. And this is true no matter how much you paid for your speakers.
The example given above of a 19x37x10’ room has a 38 distinct resonances below 100 Hz [Axials=10, Tangentials=20, Obliques=8]. That’s 38 frequency zones supported by the natural reverb field of the room… alternatively an 11x12x8’ room has only 8 [Axials=4, Tangentials=3, Obliques=1]. In the smaller room you’ll never get your sound system to have anything approaching the smooth low frequency response it will exhibit in the larger room absent acoustic treatment. Therefore, it is rational to budget accordingly.
