Ask HN: What's the risk of total splash page change?

I run a site with good inbound SEO at present, but we can clearly 20-30x traffic in our niche based on position. What seems to hold us back is the terribly massive JS size of this web app. The problem is, React.lazy and other tools be damned, it can't be changed to load the '/' page in a reasonable time. Too many network calls to check if logged in, firebase imports, graphing libraries, css libraries, drag'n'drop and other stuff is used in code present due to a demo or authentication part presently used on the home page.

I am working on hitting Googles Web Core Vitals / LCP requirements and realised our simplest solution is the best: Jump in the deep end. Switch to using a html-css only home page with images, and an iframe below the fold for the interactive demo.

What is this move likely to do? If we keep ~60% of the text content, could we see our SEO completely tank or jump up, or nothing? How long might be wait to see a change? Should we do this before or after a seasonal peak of traffic? I really value any experience with doing this. I feel we need to do it. If it's just our mb of code causing the issue in ranking (actual app is fine and liked by users), I feel like we are missing out on so much organic growth. We deliver an authentic and industry-first app that has good time-on-site for those that survive loading the site.

Any experiences or advice on this?