Is corporate engineering a dead end?

Yes, I think it's possible, but likely harder; it will be especially hard if the internal tool you support doesn't already have a huge number of users or it's not very critical to the running of the company.

I worked on Google's Engineering Productivity (EngProd) org for a bit, whose mission is to improve the productivity of product engineers via better tooling and test infrastructure support.

EngProd definitely had room for L6+ engineers (most were managers), but they were rare, and were really only needed when the product engineering org they supported grew beyond a certain size.

Fundamentally, SWEs at big tech are usually evaluated similarly across the company. That means especially at L6+ levels, engineers need to find big areas of responsibility and do work that result in large impact. Doing this for an externally facing product (hopefully a "core" product of the company), compared to internal tools, is vastly easier, because you can usually find some correlation between the work you are doing and the bottom line of the company. For example, for a product engineer to justify their impact, they can say did project X that resulting in improvement Y for Z amount of users, which resulted in XYZ amount of additional revenue for the company; in contrast, an internal tool engineers would certainly have a smaller set of users and murky revenue impact numbers to go off of at best. As such, career growth on these external product tend to be smoother, though YMMV.

So, if you only look at career advancement perspective, externally facing product teams are better. But on the flip side, work life balance on internal teams are often really good and tend to be more stable.