I haven't explored the methodology info yet to attempt to figure out why there's such a big discrepancy, but figured I would point it out here since you may already know the answer without needing to do any digging to find it (maybe there's a big difference in the make-up of the "total stock market" portfolios across the two platforms?).
Excellent point. I'm certainly not an expert on cFIREsim, but I'm aware of a few differences. Most notably, the total stock market on PortfolioCharts is basically VTSMX that includes mid and small cap companies, while cFIREsim is based on the Shiller data which tracks the S&P 500. The extra small cap weighting in VTSMX boosts returns a little. For a more direct comparison, use Large Cap Blend. Beyond that, we appear to also use different inflation data, and it wouldn't surprise me if there were a few differences in calculation methodologies. Without seeing under the hood, it's hard to know for sure. Small differences in data can make a noticeable difference 30 years out.
FWIW, I think cFIREsim is a great tool, too. I'd recommend using both to get a flavor of retirement scenarios from a few different perspectives.
EDIT -- Researching your comment also helped me find a bug with the expense ratio calculation. Thanks!