Platforms are evaluated on feature lists and paid for in operations. That gap is why so many stores end up on the wrong one.
A feature grid tells you a platform supports something. It does not tell you whether that support is native or an app, and the difference is the whole cost question. Native means it is included, maintained by the vendor, and it keeps working when everything else updates. An app means a monthly fee that scales with your order volume, a second vendor who can raise a price or stop maintaining it, and one more thing that can break in an upgrade.
So the useful audit before choosing is not which features exist. It is: list what your store actually has to do, mark each one native or app, and add up the second column. That number is your real monthly cost, and it is almost never the number on the pricing page.
The other half is where the ceilings sit. Every hosted platform has limits on variants per product, categories, API calls, and how far the checkout can be changed. None of them matter until your catalog reaches them, and then they all matter at once.
If the platform is not settled yet, the decision above this one is e-commerce web design, which argues the build from catalog, checkout and operations rather than from a vendor. Where the storefront turns out to need building rather than configuring, that is website development. The rest of what we take on is under our services.