There is no apparent reason why companies cannot build tokenized services directly on top of established tokens such as Bitcoin or Ether. So, why bother with tokens?
Adding tokens to the system saddles users with enormous inconvenience and in fact likely hurts more than it helps. It is the equivalent of, say, toting around dollars, pesos, pounds, yen, renminbi in your wallet, and then having to fork out the appropriate currency and change at every merchant you visit. Clearly a disincentive for general use or mass adoption.
The ideal form of money means we already have the token that we need. Instead of buying a token that’s useful for a specific use case, the ideal money token can be used at any time and any place. Since markets for universally-accepted money will always outpace the demand and market for any single service, fixed supply utility tokens will always encumber their users with the risks of higher volatility and lack of liquidity.
The Bitcoin token has value because it directly represents the asset it tokenizes: an unforgeable, costly and decentralized ledger. The fact that we are willing to pay more than the intrinsic value is reflected in Bitcoin’s monetary premium. The Bitcoin token-asset relationship is closed-loop. There is a direct, self-enforceable linkage between the token and the asset. The token doesn’t reference any input outside of the Bitcoin ecosystem.
This is typically not the case, however, for utility tokens because they all reference external inputs via oracles or other types of centralized solutions. They are open-loop. As such these tokens cannot be worth as much as the assets they try to represent, and even then, only up to their weakest link; how asset ownership is actually enforced. This huge built-in liability must be discounted from the value of the tokens.
In the case of a utility token with a variable supply, this represents a fixed unit of processing power, vpn access or any other service. In fact a stable coin is created that represents the fair price of the represented unit. In the best-case this encumbers the utility token with a useless layer of indirection. Further, any utility token will inevitably have an exchange rate with Bitcoin which in turn will be worth at most the equivalent to its service that is payable in Bitcoin.
This can be compared to pre-paid access keys to services. As the access key has no value in and of itself, it can only have value if the thing it protects has value. In the case of company-issued tokens, their value hinges on the company actually providing a service and making good on their promises. In the worst-case the company can shut down their service, disappear tomorrow, and your paid keys rendered worthless.
Of course this could happen to Bitcoin too, but such a scenario would require the entire Bitcoin network to shut down, a much more difficult proposition.
There is an argument that tokens can be used in decentralized smart contracts that are trustless and therefore give the token value. Unfortunately for most cases, if not all, these contracts eventually will have to interact with centralized solutions.
So, considering all the inconveniences and a user interface that is cumbersome, let’s be candid . . . at the end of the day the above approach enriches the founders and early adopters but leaves the rest of us with a user experience that is a pain. Of course you could still use tokens. But why bother?