Anyone tried webaccelerator.google.com?

For anyone who missed the beta: a short, practical checklist to evaluate any client-side "web accelerator" and decide whether to try it. This responds to the thread comments from @jewboy, @Dani, @Antonbomb22 and @seobook by turning their concerns into reproducible tests and concrete checks.

Start with repeatable performance measurements. Use the browser Network panel (disable cache) and capture median values for TTFB, DOMContentLoaded, and total load size across several loads. For a quick command-line timing snapshot, run:

curl -o /dev/null -s -w "namelookup: %{time_namelookup}s\nconnect: %{time_connect}s\nstarttransfer: %{time_starttransfer}s\ntotal: %{time_total}s\n" http://example.com/

Install the accelerator, repeat the same tests on the same pages and network, and compare the numbers. Also run traceroute/ping to spot path changes.

Check privacy and server-side effects. Read the product privacy policy before installing. On your webserver, watch logs for unusual patterns after install: new bulk requests from single IP ranges, added headers such as Via or X-Forwarded-For, or changes in User-Agent/referrer that could skew analytics. If you can, test a small private page that returns a unique string to see whether requests are proxied or cached.

Security and SEO notes. Inspect HTTPS certificates in the browser before and after installation to detect any TLS interception. Client-side accelerators usually do not change how search engines index your site, but they can alter real-user metrics and server logs — so interpret analytics carefully.

If results are unclear or the privacy policy is vague, wait for independent reviews or uninstall the beta. Basic uninstall via the OS application manager and re-running the tests will reveal whether behavior returns to baseline.