Yes—the key difference is not just “more colors.” In Photoshop-style HDR, 32-bit images use floating-point values per channel, while 16-bit and 8-bit images are typically integer-based. That gives 32-bit HDR vastly more usable dynamic range.

So when you merge exposures into a true HDR image, 32-bit can store the expanded scene range without forcing it back into a limited output range. A 16-bit file generally cannot hold that same full HDR range, so to save into 16-bit you must either tone map or clip values—both are forms of data loss.

That’s why Photoshop separates the steps: first create the high-range composite, then tone map it later for normal display or printing.

Use 32-bit when your goal is to preserve the full merged dynamic range and keep maximum flexibility for later editing and tone mapping. Use 16-bit after you’ve decided how to compress that range into a standard image, while still keeping smoother gradients and better editing headroom than 8-bit.

In short: 32-bit is for storing true HDR data; 16-bit is for a tone-mapped or otherwise reduced-range result.