I guess it matters a lot what happens to the grown out wood as well. If it's burned after cutting it down, you're back to zero for that particular tree. So assuming you're not storing large amounts of cut-down wood, the relevant metric is amount of bound carbon per surface area, and I think larger trees might have a higher density (in the long run) than smaller ones.

Coastal redwood lives for a looong time, up to 3000 years. All that carbon is eventually going to go back to the atmosphere one way or another, but the time scale looks long enough to keep some carbon safely sequestered until after we've figured out how to live in harmony with our planet -- assuming we don't manage to kill all the trees in the meantime.

No - a large tree grows at lower rates, but puts on much more mass (sequesters much more carbon) per year.

edit: see figure 6 in https://journals.plos.org/plosone/article?id=10.1371/journal..., for example. And http://science.time.com/2014/01/15/study-shows-older-trees-a... (https://www.nature.com/articles/nature12914) specifically for redwoods. Key quote: "at the extreme, a single big tree can add the same amount of carbon to the forest within a year as is contained in an entire mid-sized tree"

But clones are not the right way, because genetic variability is zero and is easy to obtain and germinate seeds that make a better root system.