a steel man for "token centric" bridging:

to date, both users and bridge operators have had a strong preference for native assets vs. receiving a bridge token.

I think preferences might actually change.  3 things to keep an eye on:

1) v simple observation but all tokens on L2's are bridge tokens.  they just rely on the canonical messaging bridge vs. a third party one.

it seems pretty likely that the lines between, for example, SuperchainERC20's and OFTs blur... a lot

3) a lot of interchain tokens are actually attracting DeFi liquidity.  this changes the game for bridging (& solvers) & it's just generally a good thing from a pure interop pov.  success begets success, look for teams to copy the OFT vanguard.  wormhole, axelar, hyperlane, lz -- all just doing a lot of good & interesting things here.

2) from a UX POV, bridge tokens suck because there's no where to use them.  you can't pay gas, you can't swap to something else, you don't want to hold it, it's just this random IOU.

users may increasingly want to hold these tokens if they:
i. have new features, like native yield
ii. a lot of new chains have native AA.  if you can pay network fees in any token, why not a bridge one.  it doesn't seem crazy for paymasters to basically be doing the rebalancing.  that actually seems like a good idea.