Alvin Smith :
Sounds like fun but, with an 18 month product cycle (per major platform), i'd recommend the compute core be purchased, all at once.
As long as YOUR full system build is NOT complete, well within 18 months, you are flat out losing cash. Sometimes things go on sale because something ain't right, with them ... best to test ... quickly.
Yes and no. Motherboard, CPU, yes. Cases,RAM ,PSU, DVD drives, not so much. Like I said, it helps to know what your doing.
But take ram as an example. I realized last summer that RAM prices were as low as they would be for a while and I purchased an 4 GB set that I wasn't quite ready to use, but expected to soon. Use in a month turned into lying in a corner for a year. Into a my new build it went. Today that $40 RAM goes for $100 I actually saved $60. Not a plan, but it worked.
RAM will be going up ... 'till at least next April 1st ... you might want to risk a "good deal" on a blind RAM pre-purchase.
I wouldn't buy RAM right now unless you had immediate use for it.
The economics are really quite simple ... A computer completely depreciates to ZERO, well within 5 years.....
Your lumping all components into one category, but not every component depreciates that quickly. That system I just built, all components used except the motherboard, cpu were available 3 1/2 years ago at the same price. Zero depreciation in 3 1/2 years. Cuts into your argument a bit.
I mean ... there is ALWAYS a "counter argument" or an "alternative philosophy" to any stance ... This is clear cut, tho ... I have won this debate (where sanity may rule).
We are obligated, by the rules of this forum, to give our best advice (or at least not to give bad advice, intentionally) ... Gonna have to recommend you disregard the last poster.
You won this debate? Anybody who disagrees with your conclusion is insane? Maybe some people are just smarter than you are Alvin Smith.
Like I said, I don't really plan it. But I have had parts sit around for 3+ years and always came out ahead and usually way ahead. It is a solid proof of concept that a skilled person could build a system over time and save a lot of money doing it.
Of course, you have to be really good to make it work. I guess Alvin Smith is not.