Really depends on what it is -- CVCV.nets are doing quite well in example now that none are available, yet Quad Premium LLLL.nets aren't doing all that great.
I would make sure you have the financial capacity to hold whatever buyout you're creating for at least 2 years in the event that it takes off considerably slower than you had initially envisioned.
Let's say I've found that all ##DN.com's are gone, except for 100 that are left. So DN stands for domain name, so would it make sense to register the register the remaining 100 and in theory, would it push up the value of any ending in DN?
So then I'd potentially, be able to double my money instantly, well, not double, but get more back?
Depends on the demand in the market. If demand is created for such names in the future you can bet they will be worth something. This is why we are in an industry full of speculation.
.com would have been a sure bet, that's for sure. If we look to .com, we see that good acronym combinations like DN, CA, RX, UK, etc fetch very noticeable premiums over comparable domains not containing those acronyms.
I really don't know if this will repeat itself in .net or other extensions in the 4 letter space -- it would certainly be a long term bet imho. It would be best to stick to what's working in those markets -- getting quad premium LLLL.nets with "DN" in them for the price of regular quad premium LLLL.nets in example is one way you can be assured to come out ahead.
Before you go regging 100 + domains you should ask yourself: is there a market? could there be a market? Will demand for such names rise in the future? Will this be long term? Short term? Development plans for some of them? Why are they not regged now?
I think reece is right on this one. It will prolly be a long term investment.