Gold has a capped supply. It’s just unknown. So it keeps inflating because inflation always refers to actual supply in the here and now and not to imaginary supply in the future. 

Bitcoin could have very well failed in its early years. Would it then have been inflationary or deflationary?

If future capped supply makes a money deflationary, how do you measure or even call future increase in deflation by lost coinage? 

The definitions fit. They have to be applied correctly though.