Could a small Mars Cycler be used for repeated sample and big data transfers?

I'm not sure data return in a physical format is worthwhile. Samples are another question, of course.

If we're already willing to wait seven years for it all to come in, why not transmit data slowly throughout that period via a relay satellite? This means you can start interpreting it early (and help inform the ongoing sample collection work); you avoid the "all your eggs in one basket" problem; you escape the risk of physical data corruption; and, best of all, you get flexibility. You can relay information from other orbiters, from other (non-sample-return) vehicles, etc.

But surely a relay satellite is too slow? Well, not really. Conveniently, we have a figure for the amount of data returned by the Mars Reconnaissance Orbiter over its first seven and a bit years; over 200 terabits (25 TB). It's reasonable to assume that a dedicated relay satellite would have the ability to exceed this, and if we can transfer dozens of TB every seven years, we're seriously competing with the physical storage approach.

Only a small fraction of the data sent through MRO is relayed from the surface, but if we knew there would be a dedicated and high-capacity relay satellite, rovers and landers would no doubt be built to rely on that infrastructure, and high-bandwidth relay uplinks on them could be prioritised.

Yes, we'd still have to keep the data relay satellite working. But we're doing pretty well with Mars orbiters - every orbiter that's made it to Mars in the past fifteen years is still going - and if we built a cycler, we'd have to worry about keeping that going.

All told, from the data transfer perspective, it's probably unnecessary.