I have indeed received your request, however, έκαμα is a valid past form for κάνω. I chose to put it in that form as well because I find to be helpful in distinguishing between the aorist and the imperfective past tenses (it's έκαμα and έκανα instead of both being the same).
The autosuggest is actually making a lot of requests to the database and with the current resources I have, I can see why it might be slow. Sorry... And I have fixed the English spelling mistake(s) (it also applied to that in the word "go") you mentioned, thanks!
I definitely see the need you mentioned for identifying the verb and I am aware of NeuroLingo (I used it myself a year ago when I was learning Greek actively). I have given a looooooot of thought to this myself. It is, however, a bit hard to implement: the verb forms often follow different rules so that it's hard to trace them back and there are a lot of exceptions thus it would take a massive amount of programming work for me to get this working properly. A bit too much, at the present moment.
I could also just store all of the forms in the database and just run a search function through the database to find the proper form but that would take not only having a lot of words added (we're working on it) but also a lot of additional database work which would make the service slow.
That's why I decided to focus on keeping it as it is and on word growth at present. If it will show signs of taking off, I will definitely be taking a second look at making it able to recognize any word form fully. After all, that would be cool.