Available quota different then advertised - gemini-2.5-pro model and gemini-cli

So I am opening this thread because the service being advertised and the service being delivered are very different. I am a subscriber of the premium developer program, available in https://developers.google.com/program/plans-and-pricing and it states that the quota for Gemini-cli is: 120 requests per user per min
1500 requests per user per day
(shared with Gemini Code Assist)

. Now I want to raise 2 points:

1 . gemini-3 models are unusable. I have tried to use it for weeks now, and the service is always unavailable or busy.
2. the quota for the models offered is different than whats advertised. Proof is that I have not used it at all during the weekend, started using it today (monday morning) when I get prompted with the message that I reached my quota. When running /stats, got the information that I issued 253 requests total, which is 20% of what is advertised when you subscribe to the membership program.

3. The entire day I had to wait 5-10 minutes and keep retrying because I was getting the message “model is busy, do you want to cancel or keep trying?” AT LEAST once per interaction.

Is anyone else experiencing this? I feel like I am getting a lot less then what I paid for, not only quantity wise but also quality wise

It happened agian today, but now instaed of the 253 requests, my quota ended and the /stats window shows this:

So I keep wondering if I will get a refund for the difference in the limits advertised/delivered