Inflation:
First year: 100%
Second year: 50%
Third Year: 33%
Fourth Year: 25%
Fifth Year: 20%
Sixth Year: 17%
Seventh Year: 14%
Eight Year: 12.5%
Ninth Year: 11%
Tenth Year: 10%
...
Year 30: 3%
Year 100: 1%
As you can see, even without halving of block rewards, and without a fixed max number of coins, inflation naturally tapers off asymptotically. Would this potentially better help miners, reduce the chances of deflation (or at least make the deflation basically minute), smooth the huge price increases that happen during the early years as people adopt the coin, reduce the huge 'unfair' advantage that some people complain Early Adopters have, and generally help Bitcoin?
Not saying I'm in favour of changing the protocol, it would be way too big of a change to make sense, and for a fairly small potential benefit, but still its a fairly interesting question.
I think the block reward halving was responsible for making Bitcoin much better. Lower costs for savers in the earlier years. Now less BTC printing to pay for the network but enough to make it secure.