@Joseph Rodriguez yeah you're reading that right — the 1 second figure is a hot start number, not cold start. The 26-32 seconds someone mentioned above is pretty consistent with what I've seen too, and that's actually pretty normal for the chipset class — open sky, good signal, no almanac decay. If you've been sitting idle for weeks you might push toward the 45-60 second range depending on how stale the ephemeris gets, though in practice most modern receivers handle that gracefully once they're getting signal.
On the flash/battery-backed SRAM question — I don't have a definitive answer off the top of my head and I don't want to guess on something that directly affects your firmware design. That's genuinely worth getting clarity on from Moamen or through the Dev Hub resources, because if it does retain almanac across full power-off that changes your buffer logic significantly. Worth opening a direct support thread if the docs don't spell it out.
For the 18Hz and TTFF interaction — the fix rate doesn't meaningfully impact how long the initial acquisition takes. TTFF is a satellite search and signal lock problem, the output rate only matters after you have a fix. So you're not paying any penalty there.
Given your use case I'd design conservatively — assume cold start every boot, buffer for at least 45-60 seconds before you trust a fix for logging, and maybe add a quality threshold check on HDOP or fix type rather than just time elapsed. That way your logger is robust regardless of how long it's been sitting.
