On `Universe` and `ObjectStore`: Is it a good practice to run the `UniverseSelectionModel` on a weekly basis, save the resulting symbols to the `ObjectStore`, and then load these symbols from the `ObjectStore` on subsequent days to speed up the algorithm’s initialization?
2. On `ATR` Indicator Resolution: My universe selection provides data at `Resolution.DAILY`, but I have registered an `ATR` indicator for these symbols at `Resolution.MINUTE`. Does QuantConnect automatically handle this resolution mismatch by warming up the necessary minute data for the daily symbols, or will this approach lead to errors or inaccurate indicator values?
3. On `BenzingaNews` Timestamps: For processing news events in real-time as they are released, which property of the `BenzingaNews` object should be used for the most accurate timestamp? Is `news.end_time` the correct property to compare against the current algorithm time (`self.time`)?
4. On `self.notify.telegram` Reliability: For a live trading algorithm that may generate a high volume of alerts, is the built-in `self.notify.telegram()` method reliable enough, or is it best practice to implement a custom queuing system with rate-limiting to avoid any potential issues with Telegram’s API limits?
5. On `SetWarmup` with Dynamic Universes: How does `set_warmup()` interact with a `UniverseSelectionModel`? If I set a warmup period, is that warmup logic applied to new securities that are added to the universe by the selection model after the initial algorithm warmup has already completed?