My suggestions would be lets try testnet3, get safecoin 1.0 implemented and run some tests for divisions if required. I would think that if safecoin went ballistic then it can become like a dollar and we can create subdivisions similar to dimes. That way we do not renege on the 1:1 split and if this is massively successful we will also have a significantly larger dev team to handle changes. We are currently really busy and making very large but effective changes along the testnets (but slower than I would like so far 
) and this is potentially an issue that can be tackled along that route.
As a small dev team we need to consider security and code quality for a system that when you consider it is all levels from level 3 to 7 in the 7 layer model. It’s already a large body of work and if it seems we need to increase supply then I feel it should be part of a much larger conversation.
The knowledge that we cannot tell the future is valid and I find the edge effects of fast decisions are always considerable. This is not really a decision that can be made on the fly without considerable thought and consideration for all the effects (space usage, size of coin space, churn implications etc.). More importantly though I think that if a safecoin was very valuable a sub currency would appear so instead of 0.x safecoin there would be a subdivision of 1 bit say (doubling the number of the sub currency etc.). This is what I mean though we are then dealing with inflation of a currency by not printing money in a way (or pre print a ton of it at the start).
In the network things like buying space / computation would not be per byte or cpu cycle but large chunks (let me store for a year without worry for a few safecoin as a normal user kinda thing). Many other things will be similar. Of course the greatest value will likely be off network purchases and transfer of value for goods and services, and where the issues will appear first.
If there is is an issue that a a safecoin is worth say $0.50 and it becomes a problem I am sure we can overcome it with subdivisions as planned and that is fair and balanced and initial holders are rightly rewarded without doing maths in their head to figure what they bought. The subdivisions should be well thought out and named properly though and this is where we win. Bitcoin nearly did it with bits but by the time it happened people were already confused. If we need smaller currency units probably the first thing to do would be to decide on the unit and the name and release it properly and to a wide audience. Unlike bitcoin we will not already have .000X ability which I think is good as it will prevent confusion.
We can do this 2^32 times if needed (it won’t be) in due course and really well planned. I think any guessing of the future is futile, but being able to cope with it is effective and something we can do. When I say I don’t like subdivisions, this is this what I mean, I don’t want to pay 0.000000X of anything, but if we do this right nobody will have to. 