Resampling algorithm and Padding

[vincentsarago]

Following discussion from #647 here is a small analysis of the impact of the resampling and padding

https://gist.github.com/vincentsarago/ff73ea5dbf128949d951851cc8454b4f

Tile within the dataset zoom range:

☝️ There isn't any highly visible edge artefacts but we know the data is somehow slightly different (but nothing visually noticeable with or without padding)

Overzoom (tile at zoom > than dataset max resolution)

☝️ We can clearly see edges artefact and disappear when using padding=2. Also noticing that padding has no influence when using nearest resampling.

the TLDR, padding is useful mostly when overzooming. When in the dataset's zoom range there is not highly visible edges so padding do not have huge impact.