A screenshot proves that an image existed. A shared statement link proves a little more, and a lot less than it appears to: it can be a selected account among many, a demo, a period chosen after the fact, or a record that stops at the moment the curve stopped going up.
None of that requires dishonesty. A trader with four accounts who shares the best one is doing what any of us would do. But an allocator has no way to tell that apart from a trader with one account and four good years, and the safest assumption is the unflattering one.
Survivorship is the quiet problem
The accounts that get shared are the accounts that survived. The blown ones are not in the folder. An allocator reading a stack of shared statements is reading a filtered sample and knows it, which is why the stack rarely gets read at all.
What replaces it
Measurement at the account level, from the source, continuously, across every account a trader connects. The result is less flattering than a screenshot and far more useful: drawdowns stay in the record, quiet periods stay in the record, and the number at the end means something because nothing was removed to produce it.