Site Reviews (chummytees.com)

We're looking to improve the helpfulness of our customer review page(https://chummytees.com/pages/reviews) as well as the review sections on each t-shirt product page. We're currently showing the item purchased, the style/size/color, a photo of the product, the customer's name, rating, and the review itself. All of our 10,000+ reviews on the site are from existing customers and are collected after purchase via email using the stamped.io shopify app. Since we cannot reasonably ask 10,000 people to update their reviews with our new format, we would like to utilize word-searching functionality for existing reviews and implement new features for future review requests.

We would like to be able to further help customers by adding the following functionality:

1.) Adding the ability to see(by using a selector/dropdown) all reviews for a particular style (e.g. the womens v-neck), as these are much more helpful for someone wanting a womens v-neck than seeing reviews for a toddler tee, mens tank, etc.
2.) When a review is requested, the form would be updated to include a sliding scale they could select for the sizing of the t-shirts they received, going from “Runs Small” to “True to Size” to “Runs Large”. With the aggregate information, we could provide to the customer something like “80% consider this true, 10% Runs small, 10% Runs Large" for any specific style. We are not entirely sure how to request this additional data in a structured way since the app only has specific fields. We have used these other review apps in the past(https://www.reviews.io/company-reviews/store/chummytees.com) and (https://loox.io/s/VJZgFVYaGF) and would be willing to switch to a different code if required, but only if it integrates with Google shopping. 
3.) For the existing 10,000+ reviews, we'd like to be able to utilize word searching functionality for point #2. For example, if the review says “true to size”, “runs small”, “too big”, “too small”, and other variations, it would be considered in the calculating of the prior point. 

These are what we're hoping to implement if feasible. If there are other things we should consider for an apparel brand, we're open to ideas as well.