Well, if you assume economic growth, a fixed number of counting tokens will have to represent larger values over time.

The real trick is to read 'deflationary' as 'will create deflationary pressure', not as 'will inevitably cause deflation at every moment in time'.

You mean, assuming there was only one Bitcoin: if a VW Beetle costs 1 BTC this year, and next year a Porsche is published (growing the economy by replacing cars with better cars), the Porsche would also have to cost 1 BTC because there isn't more than 1 BTC?

That seems incorrect. While impractical, there might be other solutions, for example the Porsche could cost 1 BTC+ 1 Carrot, or maybe it could cost 2 BTC anyway. The buyer just wouldn't be able to pay in one go.

Marginal effects don't really reduce well to narrow examples like that.

What I mean is that if you assume a healthy trade in something of limited abundance, you should be prepared for demand to drive up the value of that thing. It might not, but given the simple dynamics of supply and demand (microeconomics isn't really controversial), it is prudent to be aware of the possibility.

(I do sort of assume that people are mostly using bitcoins to complete transactions and speculate, I don't assume there is a healthy trade in them)