Well, you can't sell it as both. There's only one way: digital gold. Big blocks kill the digital gold feature, which kill the digital cash feature too (you can have digital cash when digital gold is not an option), since big blocks centralize the network.
With conservative blocks, you can get digital gold, and you can build on top a second layer, which is debatable if it would be considered digital cash or not, but it's definitely more private than traditional payment systems and fiat currencies.
I agree somewhat on your opinion about (too) big blocks, but your conclusion, for me, is wrong.
It's actually the other way around: Bitcoin can only be digital gold if it's also used as a currency ("digital cash"). It is not that important how it's used (e.g. if used in supermarkets or mainly for transactions) but it should cover as many use cases as possible. (Micropayments, for example, is not really an important on-chain use case and for this case I think LN is better).
The reason is that for the "digital gold" feature Bitcoin needs at least a certain level of value/price stability. Stability can come from usage (as a currency) and liquidity. If there is no usage and liquidity its value will be vulnerable to speculation attacks, because Bitcoin itself is not "backed" by anything but its users themselves. The more users and the more usage, the more stable it can be.
What to do, then, if bigger blocks are not desired? There is no single answer, but LN, pegged sidechains, extension blocks and on-chain density improvements (without too big blocks) are as well needed as a gradual and conservative block size increase (17% per year as in BIP 103 is a good orientation point although I don't like the automatism behind this proposal).
PS: I voted for "Both" and although I don't agree to the full post of the OP, I think this phrase is worth a quote:
Bitcoin's value depends mainly on how many people can safely use it, and if it's not in use by the largest number of people possible, it will not achieve it's full potential/value.