If it only took 1-10 billion years then that would be a problem. Splitting the work between 1-10 billion computers would then make is a 1 year quest only. How you know how ASICs compare to everyday computers regarding the mining speed. Imagine an ASIC which sole purpose it is to find a given bitcoin address. That would probably reduce it to 10k-100k of these machines (plus you can add in some normal computers to substitute for not having as many ASICs in the beginning at a 1:100 rate (for the beginning, these ASICs would, over time, be 1000 times and more faster than a normal PC).

Now also imagine you don't only look for only one address, but... lets say the Top 1000 BTC addresses. If this 1-10 billion years and there fore the 1 years with the power of 1-10 billion machines is correct and the worst case scenario, hat would leave you with ~2,74 of those top 1000 address' private key being found PER DAY. If it's not the worst case scenario but an average, that would leave you with only ~1,37 private keys per day.

Someone please enlighten me if the 1-10 billion thingy is correct because that could be a huge problem regarding the fact that computer power increases rapidly.