Bitcoin is transacted immediately, like a creditcard or debt.
For bitcoins to receive a confirmation, thus creating a final settlement, it must be within a block.
For Bitcoin to have faster confirmation times, there are three possible choices:
1. Shorten the block interval
2. Raise the max blocksize
3. Create second layers designed for faster and more tps.
Those three choices is what the current "blocksize debate" is about.
Each choice creates other problems that can or can not be currently solved.
Some of the opinions about each one, are also in debate and are just opinions.
No one knows anything for sure and an incorrect choice could damage the system and its future.
number one of the list is not a viable option..
it has many repercussions and changes many dynamics of bitcoin. its not part of the main debate. just scenario opinions and personal chitchat
number 2 can work once to stick ductape on the doomsday guys that throw randomly large numbers into the air that have not realistic rational reason to be concerned with in the next few years.
number 3 can work but more for the niche of day traders, facuet raiders, gamblers. not so useful for people that only buy one item a day/week/month.
also number 3's 'settlement' time becomes longer than 10 minutes. but atleast the 'trust' of instant acceptance is higher
funny part is by adding more double spend risk features like RBF and CPFP it has made 2017 more risky to accept a zero confirm.. not less risky. thus all that promise of segwit solving double spend risk has fallen flat