Members of the computer music community argue continually about whether it's better to opt for the value-for-money of a powerful PC system or for a more user-friendly but costlier Macintosh. I've no doubt that these discussions will continue regardless, but the reality is that either system provides more than enough horsepower for recording music. There are very few hit records or classic albums that couldn't have been made using 24 tracks or less, and where more tracks were used, it's often the case that the drum kit and unused vocal takes accounted for over half of them. These days, with virtual tracks at our disposal, alternative takes cost us nothing. What's more, while keyboard parts were once played and recorded to tape like any other instrument, in today's computer-assisted studio they tend to be handled via MIDI, which has a very low processor overhead.
Of course, you do need computing power to run virtual instruments and plug-ins, but again, if you analyse most classic records, you'll find that there isn't as much going on as you might expect. Even dance mixes can be broken down into a modest number of simultaneous tracks, many of which are sampler-based, so it can be argued that processer speed isn't a serious limitation for most users, provided they are putting their songs together in a sensible way.
I operated an eight-track tape-based studio back when eight-track was the pinnacle of home recording technology, and I recall recording a band using only seven of the available tracks. Admittedly, it was a small band, but the point is that once they discovered there was a track free, they agonised over what else they could put on it, eventually settling for tambourine. The song wasn't written with a tambourine part, and it didn't really need one, but because there was an unused track, they felt they had to put something on it.
The urge to use every available track is still prevalent, and current recording systems, both software and hardware, provide enough power in this respect for full-scale musical megalomania. Sure enough, a common fault with many of the home recordings I hear is that there's no space left in the mix. At one time this might have been down to the use of too much reverb, but nowadays it's more likely to be due to too many unnecessary layers. Part of the problem is that many of us use a lot of off-the-shelf samples and synth patches, and if we can't find something exactly right, we try to create it by layering multiple elements. While this can be effective for string pads and the like, it's more likely to have a muddying effect, leaving you with a mix like an overworked watercolour painting.
Perhaps a worse mistake still is to use excessive processing to try to knock an audio recording into shape. If the unprocessed recording doesn't sound pretty good in the first place, then it's probably not good enough to use. Examine both the performance and your recording technique and see how you can get closer to the sound you want. Often this is as simple as moving the microphone or hanging up another duvet. Similarly, choose your synth sounds and samples with care and edit or EQ them where necessary to make them sit properly in the song. Not only will this reduce the load on your computer, it will also almost certainly result in a better-sounding recording with more space and definition. This might sound like hard work, but it's not nearly as difficult as thinking up things to record on all those spare tracks — believe me!