I think market sizing is a fairly valuable activity. Worst case you are forced to think about it and write down assumptions you can test. The initial guesstimate might as well be throwing a dart but you can use it as an orientation to see what direction you are way off in.
Even if you feel like you're creating a new market segment (blue ocean) it can't hurt to think through it. What might the structure look like, what other market segment could be similar etc. etc.
At the end of the day you're thinking about aggregated customers. Even if you think market sizing is a waste of time you still need to think about reaching customers and how they interact with one another (feedback loops, word of mouth).
Even if you feel like you're creating a new market segment (blue ocean) it can't hurt to think through it. What might the structure look like, what other market segment could be similar etc. etc.
At the end of the day you're thinking about aggregated customers. Even if you think market sizing is a waste of time you still need to think about reaching customers and how they interact with one another (feedback loops, word of mouth).