The actual starting stack sizes -- and the actual starting blind amounts -- are far less important than a) the number of starting blinds in a starting stack, and b) how quickly that number of starting stack blinds is decimated by the blind structure.
A T200 starting stack with 1/2 opening blinds is 100bb per starting stack, the exact same as a T10k starting stack with 50/100 opening blinds... or T1000 stacks with 5/10 opening blinds. All 100bb events -- regardless of how constructed -- will play exactly the same, provided the blind structure percentage increases and the blind level times are the same for each event.
Some starting configurations are more efficient than others, however. A T25-base set is not one of those that qualify as efficient, unless one chooses to use T500 & T2000 denominations and exclude T1000 chips.
A T.25-base set actually works best (T.25, T1, T5, T25) from an efficiency perspective. A T1-base set (T1, T5, T25, T100) works, but at a 5:1 replacement ratio for the lowest two denominations, is inherently less efficient (or cost effective). A T5-base (T5, T25, T100, T500) structure runs into issues once the field size reaches certain numbers, and a T25-base set has inherent problems right out of the gate.
Most casino tournaments have historically been either T1 or T5 base, until the relatively recent (2005-ish) popularity of the WSOP (which featured a $10000 buy-in for T10000 in chips with 25/50 opening blinds). Many casino tournaments have since changed to T25 base sets and events, but don't be misled that it is better in some fashion -- it really isn't. It's just a cheap casino trick to make you think you're getting something "more" for your money, i.e. marketing.
I'd be more concerned with how the structure itself works.